Advisories

The read_getblocks and read_getheaders codec paths accepted block locator vectors up to approximately 65,535 entries (the generic TrustedPreallocate ceiling derived from MAX_PROTOCOL_MESSAGE_LEN), rather than the protocol-specification limit of 101 entries (matching zcashd's MAX_LOCATOR_SZ). Each entry in the locator vector triggers a per-hash chain lookup (HashMap::contains_key + RocksDB::contains_hash) in find_chain_intersection on a tokio blocking-pool thread. A single maximally-sized getblocks message occupies one blocking-pool thread for approximately 10–65ms. Under sustained load from …

The read_getblocks and read_getheaders codec paths accepted block locator vectors up to approximately 65,535 entries (the generic TrustedPreallocate ceiling derived from MAX_PROTOCOL_MESSAGE_LEN), rather than the protocol-specification limit of 101 entries (matching zcashd's MAX_LOCATOR_SZ). Each entry in the locator vector triggers a per-hash chain lookup (HashMap::contains_key + RocksDB::contains_hash) in find_chain_intersection on a tokio blocking-pool thread. A single maximally-sized getblocks message occupies one blocking-pool thread for approximately 10–65ms. Under sustained load from …

The z_listunifiedreceivers RPC handler panics when processing a structurally valid Unified Address whose Sapling receiver carries 43 bytes that fail cryptographic validation (sapling_crypto::PaymentAddress::from_bytes returns None for non-subgroup Jubjub points). The handler calls .expect("using data already decoded as valid") on the fallible result. Because Zebra's release profile sets panic = "abort", the panic terminates the entire node process, not just the RPC task.

The z_listunifiedreceivers RPC handler panics when processing a structurally valid Unified Address whose Sapling receiver carries 43 bytes that fail cryptographic validation (sapling_crypto::PaymentAddress::from_bytes returns None for non-subgroup Jubjub points). The handler calls .expect("using data already decoded as valid") on the fallible result. Because Zebra's release profile sets panic = "abort", the panic terminates the entire node process, not just the RPC task.

The mempool download pipeline's cancel_handles map retains entries for transactions whose verification times out at the outer RATE_LIMIT_DELAY (73-second) boundary. The tokio::time::error::Elapsed error carries no payload, so the transaction ID is unrecoverable and the corresponding cancel_handles entry (including the full Gossip::Tx(UnminedTx), up to ~2 MB) is never removed. Entries accumulate monotonically with no upper bound or garbage collection, leading to eventual out-of-memory process termination.

When pop_tip removes the tip block during a chain fork, stale Sapling and Orchard note commitment subtree root data is retained in the in-memory non-finalized state. When the chain subsequently finalizes, this stale data is written to the persistent RocksDB state. The corrupted subtree root history affects z_getsubtreesbyindex (used by lightwalletd for wallet synchronization) and could affect future chain verification that depends on correct subtree roots.

When pop_tip removes the tip block during a chain fork, stale Sapling and Orchard note commitment subtree root data is retained in the in-memory non-finalized state. When the chain subsequently finalizes, this stale data is written to the persistent RocksDB state. The corrupted subtree root history affects z_getsubtreesbyindex (used by lightwalletd for wallet synchronization) and could affect future chain verification that depends on correct subtree roots.

A single unauthenticated P2P peer can monopolize all 25 inbound mempool download/verification slots (MAX_INBOUND_CONCURRENCY) by advertising fake transaction IDs. While the slots are occupied, all other inbound transactions from honest peers and local RPC sendrawtransaction calls are rejected with MempoolError::FullQueue. The attacker peer is never scored for misbehavior and is not disconnected, allowing sustained denial of mempool admission.

The getblocktemplate RPC handler panics when parsing a LongPollId parameter that contains non-ASCII (multi-byte UTF-8) characters. The handler performs byte-index string slicing on the user-supplied string, which panics in Rust when a byte index falls within a multi-byte character boundary. Because Zebra's release profile sets panic = "abort", the panic terminates the entire node process.

The getblocktemplate RPC handler panics when parsing a LongPollId parameter that contains non-ASCII (multi-byte UTF-8) characters. The handler performs byte-index string slicing on the user-supplied string, which panics in Rust when a byte index falls within a multi-byte character boundary. Because Zebra's release profile sets panic = "abort", the panic terminates the entire node process.

Zebra's P2SH sigop counter uses a pure-Rust code path that short-circuits on disabled opcodes (such as OP_CODESEPARATOR), returning a partial count of zero for any sigops following the disabled opcode. The reference implementation (zcashd) correctly counts through disabled opcodes in its static sigop analysis. This produces a consensus divergence: Zebra accepts blocks that zcashd rejects when the block-wide MAX_BLOCK_SIGOPS = 20,000 threshold is crossed on one side but not the …

Zebra's P2SH sigop counter uses a pure-Rust code path that short-circuits on disabled opcodes (such as OP_CODESEPARATOR), returning a partial count of zero for any sigops following the disabled opcode. The reference implementation (zcashd) correctly counts through disabled opcodes in its static sigop analysis. This produces a consensus divergence: Zebra accepts blocks that zcashd rejects when the block-wide MAX_BLOCK_SIGOPS = 20,000 threshold is crossed on one side but not the …

Chain::push in the non-finalized state updates the transaction-location index (tx_loc_by_hash) before it runs the duplicate shielded-nullifier guard. When an invalid child block repeats a shielded transaction from its non-finalized parent, the assert_eq!(prior_pair, None, "transactions must be unique within a single chain") fires before the contextual validation that would cleanly reject the duplicate. Under Zebra's panic = "abort" release profile, this terminates the entire node process. The block should be rejected …

Chain::push in the non-finalized state updates the transaction-location index (tx_loc_by_hash) before it runs the duplicate shielded-nullifier guard. When an invalid child block repeats a shielded transaction from its non-finalized parent, the assert_eq!(prior_pair, None, "transactions must be unique within a single chain") fires before the contextual validation that would cleanly reject the duplicate. Under Zebra's panic = "abort" release profile, this terminates the entire node process. The block should be rejected …

The finalized transparent address balance writer processes all newly-created outputs (credits) before processing spent outputs (debits) within the same block. A consensus-valid block containing a long chain of same-address transparent self-spends can cause the intermediate per-address balance during the credit pass to exceed MAX_MONEY, triggering a panic in the finalized state writer. Because the triggering block is consensus-valid (zcashd accepts it), the panic recurs on restart when the node re-encounters …

The finalized transparent address balance writer processes all newly-created outputs (credits) before processing spent outputs (debits) within the same block. A consensus-valid block containing a long chain of same-address transparent self-spends can cause the intermediate per-address balance during the credit pass to exceed MAX_MONEY, triggering a panic in the finalized state writer. Because the triggering block is consensus-valid (zcashd accepts it), the panic recurs on restart when the node re-encounters …

A malicious peer can answer Zebra's outbound getblocks/FindBlocks request with a small two-hash inventory, then serve a syntactically valid block whose coinbase height is far above the victim's local tip. The AboveLookaheadHeightLimit error in the sync download pipeline triggers a global sync restart rather than being scoped to the offending peer. The peer is never scored or disconnected because the error type does not carry the advertiser address. On mainnet, …

A malicious peer can answer Zebra's outbound getblocks/FindBlocks request with a small two-hash inventory, then serve a syntactically valid block whose coinbase height is far above the victim's local tip. The AboveLookaheadHeightLimit error in the sync download pipeline triggers a global sync restart rather than being scoped to the offending peer. The peer is never scored or disconnected because the error type does not carry the advertiser address. On mainnet, …

The P2P codec's Codec::decode() method calls src.reserve(body_len + HEADER_LEN) after parsing a 24-byte protocol header, using the attacker-claimed body_len field. This reserves up to MAX_PROTOCOL_MESSAGE_LEN (~2 MiB) of virtual buffer capacity per connection before any body bytes arrive and before the handshake completes. However, BytesMut::reserve() sets virtual capacity without committing physical memory pages. The operating system does not allocate physical RAM until bytes are actually written into the buffer. Since …

The P2P codec's Codec::decode() method calls src.reserve(body_len + HEADER_LEN) after parsing a 24-byte protocol header, using the attacker-claimed body_len field. This reserves up to MAX_PROTOCOL_MESSAGE_LEN (~2 MiB) of virtual buffer capacity per connection before any body bytes arrive and before the handshake completes. However, BytesMut::reserve() sets virtual capacity without committing physical memory pages. The operating system does not allocate physical RAM until bytes are actually written into the buffer. Since …

Zebra records a block hash in non_finalized_block_write_sent_hashes when the block is sent to the write task, before contextual validation completes. If validation fails, the hash is not removed. A remote unauthenticated peer can deliver a poisoned block body that shares a header hash with a later valid canonical block. The poisoned body is rejected, but the hash remains cached. When the valid canonical block arrives, Zebra treats it as a …

Zebra records a block hash in non_finalized_block_write_sent_hashes when the block is sent to the write task, before contextual validation completes. If validation fails, the hash is not removed. A remote unauthenticated peer can deliver a poisoned block body that shares a header hash with a later valid canonical block. The poisoned body is rejected, but the hash remains cached. When the valid canonical block arrives, Zebra treats it as a …

An address normalization mismatch between the handshake path and the mempool misbehavior path causes a deterministic assertion panic when a peer connects via IPv4 to a dual-stack IPv6 listener and then triggers a mempool misbehavior penalty. The handshake path canonicalizes IPv4-mapped IPv6 addresses to plain IPv4 when storing the peer in the address book via MetaAddr::new_connected. The mempool misbehavior path forwards the raw transient socket address (IPv4-mapped IPv6 form) when …

An address normalization mismatch between the handshake path and the mempool misbehavior path causes a deterministic assertion panic when a peer connects via IPv4 to a dual-stack IPv6 listener and then triggers a mempool misbehavior penalty. The handshake path canonicalizes IPv4-mapped IPv6 addresses to plain IPv4 when storing the peer in the address book via MetaAddr::new_connected. The mempool misbehavior path forwards the raw transient socket address (IPv4-mapped IPv6 form) when …

The JWT signing key cache in TokenKeyResolver uses kid as the sole cache key without namespacing by authority. In applications with multiple JwtBearer schemes pointing to different identity providers, a key fetched for one scheme can satisfy token validation for another. Additionally, cached keys have no expiration, so rotated or revoked keys remain trusted until the application process restarts.

The JWT signing key cache in TokenKeyResolver uses kid as the sole cache key without namespacing by authority. In applications with multiple JwtBearer schemes pointing to different identity providers, a key fetched for one scheme can satisfy token validation for another. Additionally, cached keys have no expiration, so rotated or revoked keys remain trusted until the application process restarts.

The JWT signing key cache in TokenKeyResolver uses kid as the sole cache key without namespacing by authority. In applications with multiple JwtBearer schemes pointing to different identity providers, a key fetched for one scheme can satisfy token validation for another. Additionally, cached keys have no expiration, so rotated or revoked keys remain trusted until the application process restarts.

All Steeltoe actuator endpoints default to EndpointPermissions.Restricted, which is mapped to Cloud Foundry's read_basic_data permission (granted to Space Auditors and similar low-trust roles). Sensitive actuators including heap dump, environment, and thread dump do not raise this to EndpointPermissions.Full, so CF's read_sensitive_data permission flag is not enforced for those endpoints. Spring Boot's equivalent Cloud Foundry integration gates these endpoints with read_sensitive_data by default.

All Steeltoe actuator endpoints default to EndpointPermissions.Restricted, which is mapped to Cloud Foundry's read_basic_data permission (granted to Space Auditors and similar low-trust roles). Sensitive actuators including heap dump, environment, and thread dump do not raise this to EndpointPermissions.Full, so CF's read_sensitive_data permission flag is not enforced for those endpoints. Spring Boot's equivalent Cloud Foundry integration gates these endpoints with read_sensitive_data by default.

The Sanitizer component in the Environment actuator redacts configuration values by matching the configuration key name against a suffix list. The default list (password, secret, key, token, .credentials., vcap_services) does not cover the standard .NET pattern ConnectionStrings:<name> or Steeltoe Connectors' Steeltoe:Client:<type>:Default:ConnectionString. There is no value-based scrubbing, so full connection string values including embedded Password= and user:pass@host segments are returned verbatim in /actuator/env responses.

The Sanitizer component in the Environment actuator redacts configuration values by matching the configuration key name against a suffix list. The default list (password, secret, key, token, .credentials., vcap_services) does not cover the standard .NET pattern ConnectionStrings:<name> or Steeltoe Connectors' Steeltoe:Client:<type>:Default:ConnectionString. There is no value-based scrubbing, so full connection string values including embedded Password= and user:pass@host segments are returned verbatim in /actuator/env responses.

DataCenterInfo.FromJson throws ArgumentException for any name value other than "MyOwn" or "Amazon", despite the Java Eureka specification defining a third valid value: "Netflix". The exception propagates through the entire registry deserialization chain and is swallowed by the periodic cache refresh task, leaving the local service registry permanently empty or stale.

When MySQL or PostgreSQL service bindings from VCAP_SERVICES include TLS client credentials, the Connectors library writes those credentials to temporary files in Path.GetTempPath() using File.CreateText. On Linux, File.CreateText creates files with mode 0644 (world-readable) under the process umask, and the files are never deleted. The same key material is protected at mode 0400 in /proc/<pid>/environ.

Configuring encrypt:rsa:algorithm=OAEP does not enable OAEP encryption. Due to an incorrect BouncyCastle transformation string, the OAEP setting selects PKCS#1 v1.5, which is the same algorithm as the DEFAULT setting.

When Steeltoe management endpoints are configured to listen on an alternate port (Management:Endpoints:Port is configured), the middleware responsible for restricting access to the endpoints uses the Host HTTP header rather than the actual network socket port.

When Steeltoe management endpoints are configured to listen on an alternate port (Management:Endpoints:Port is configured), the middleware responsible for restricting access to the endpoints uses the Host HTTP header rather than the actual network socket port.

SimpleSAMLphp's SAML SP ACS path does not enforce the IdP selected for an SP-initiated login. If a saved SP state contains ExpectedIssuer = IdP A, but the ACS receives a valid response from IdP B, the code logs a warning and continues processing instead of rejecting the response. That behavior becomes security-relevant when combined with the response-processing rule that accepts an unsigned samlp:Response/@InResponseTo outside the signed assertion whenever the signed …

SimpleSAMLphp's HTTP-Artifact receive path can treat an unsigned embedded SAML Response as cryptographically valid for the wrong IdP. In the HTTPArtifact::receive() flow, the SOAP ArtifactResponse receives a TLS-based validator from SOAPClient::addSSLValidator(). The embedded SAML Response then receives a validator that delegates signature validation to that outer ArtifactResponse. Later, the SP validates the embedded Response against metadata selected from the embedded response issuer, not necessarily the artifact issuer. The critical issue …

SimpleSAMLphp's HTTP-Artifact receive path can treat an unsigned embedded SAML Response as cryptographically valid for the wrong IdP. In the HTTPArtifact::receive() flow, the SOAP ArtifactResponse receives a TLS-based validator from SOAPClient::addSSLValidator(). The embedded SAML Response then receives a validator that delegates signature validation to that outer ArtifactResponse. Later, the SP validates the embedded Response against metadata selected from the embedded response issuer, not necessarily the artifact issuer. The critical issue …

This library turned out to be vulnerable to Denial-of-Service attacks using XPath transforms. A mitigation has been put in place to restrict the number of transforms and to restrict transforms to only the transform-algorithms mentioned in the SAML 2.0 Core Specifications (and specifically refuse XPath transforms).

This library turned out to be vulnerable to Denial-of-Service attacks using XPath transforms. A mitigation has been put in place to restrict the number of transforms and to restrict transforms to only the transform-algorithms mentioned in the SAML 2.0 Core Specifications (and specifically refuse XPath transforms).

The inline query parameter on the browsable-share file download and on the authenticated user file download suppressed Content-Disposition: attachment, so an HTML file stored in a share or home directory could be served as text/html and execute in SFTPGo's web origin (stored XSS).

The public web-client endpoint for partial ZIP downloads of a browsable share did not correctly confine the client-supplied files entries to the shared directory. A requester able to reach a public share could read files located outside the shared directory, as long as the target's canonical path begins with the shared directory's name.

Recce OSS server deployments that expose the server to an untrusted network without authentication are vulnerable to unauthenticated SQL execution through the query run API. When Recce is configured with a DuckDB-backed project, an attacker can use DuckDB filesystem primitives to read and write files accessible to the Recce server process. The impact depends on how Recce is deployed, but may include disclosure of local files, tampering with Recce/dbt artifacts, …

Slack plugin approvals used the exec approver gate for plugin actions. In affected versions, a Slack user authorized only for exec approvals could resolve a plugin approval through the exec approver gate. This advisory is scoped to the named feature and configuration. It does not change OpenClaw's trusted-operator model: authenticated Gateway operators, installed plugins, and intentional local execution surfaces remain trusted unless a separate policy, approval, allowlist, sandbox, or auth …

On POSIX nodes, OpenClaw's system.run safe-bin checks could approve a command before shell expansion changed how the command was interpreted. A value that appeared to be a safe-bin argument could expand into additional shell words and become a file operand. This issue is limited to paired POSIX node execution through system.run with safe-bin or allowlist-style auto-approval. It is not an unauthenticated node takeover.

Marketplace runtime extension metadata could point at unscanned payloads. In affected versions, a package selected for installation by a trusted operator could redirect runtime loading toward hidden package content that was not scanned as expected. This advisory is scoped to the named feature and configuration. It does not change OpenClaw's trusted-operator model: authenticated Gateway operators, installed plugins, and intentional local execution surfaces remain trusted unless a separate policy, approval, allowlist, …

OpenClaw's browser control SSRF checks blocked direct navigation to private or loopback URLs, but some Playwright act interactions could trigger navigation after the initial check. A later browser evaluation could then read from the page reached by that action-triggered navigation. This issue is specific to browser control actions and private-network navigation policy. Browser evaluation remains an intentional trusted-operator feature when it is used on pages that policy allowed the browser …

Workspace .env could override Homebrew executable selection for skill install flows. In affected versions, a workspace .env in a repository opened by a trusted operator could override the Homebrew executable used by the install helper. This advisory is scoped to the named feature and configuration. It does not change OpenClaw's trusted-operator model: authenticated Gateway operators, installed plugins, and intentional local execution surfaces remain trusted unless a separate policy, approval, allowlist, …

In trusted-proxy Control UI mode, OpenClaw accepted a WebSocket client's declared operator scopes before those scopes were bound to a server-approved pairing or trusted-proxy authorization baseline. This issue affects trusted-proxy Control UI deployments. It does not apply to shared-secret Control UI sessions, which are treated as trusted operator sessions by design.

Trusted retry endpoint checks could match hostname prefixes. In affected versions, a retry endpoint URL chosen by lower-trust input could pass validation by using a hostname prefix that resembled a trusted host. This advisory is scoped to the named feature and configuration. It does not change OpenClaw's trusted-operator model: authenticated Gateway operators, installed plugins, and intentional local execution surfaces remain trusted unless a separate policy, approval, allowlist, sandbox, or auth …

Telegram interactive callbacks could skip commands.allowFrom. In affected versions, a Telegram user able to invoke an affected callback could mark the callback as an authorized sender before applying commands.allowFrom. This advisory is scoped to the named feature and configuration. It does not change OpenClaw's trusted-operator model: authenticated Gateway operators, installed plugins, and intentional local execution surfaces remain trusted unless a separate policy, approval, allowlist, sandbox, or auth boundary is crossed.

Slack and Zalo webhook secrets could remain active after secrets.reload. In affected versions, a caller with an old webhook secret during the stale-secret window could keep accepting the previous secret after secrets.reload. This advisory is scoped to the named feature and configuration. It does not change OpenClaw's trusted-operator model: authenticated Gateway operators, installed plugins, and intentional local execution surfaces remain trusted unless a separate policy, approval, allowlist, sandbox, or auth …

Slack allowFrom could bind to mutable display names. In affected versions, a Slack account able to change display name metadata could match a policy entry through mutable display metadata. This advisory is scoped to the named feature and configuration. It does not change OpenClaw's trusted-operator model: authenticated Gateway operators, installed plugins, and intentional local execution surfaces remain trusted unless a separate policy, approval, allowlist, sandbox, or auth boundary is crossed.

Skill Workshop apply flow could override pending approval. In affected versions, an agent tool call reaching the affected Skill Workshop apply path could set apply: true despite approvalPolicy: pending. This advisory is scoped to the named feature and configuration. It does not change OpenClaw's trusted-operator model: authenticated Gateway operators, installed plugins, and intentional local execution surfaces remain trusted unless a separate policy, approval, allowlist, sandbox, or auth boundary is crossed.

Shell wrapper argv could change between approval and execution. In affected versions, a command request using a shell wrapper form could approve one resolved argv shape and rebuild another for execution. This advisory is scoped to the named feature and configuration. It does not change OpenClaw's trusted-operator model: authenticated Gateway operators, installed plugins, and intentional local execution surfaces remain trusted unless a separate policy, approval, allowlist, sandbox, or auth boundary …

Some internal command handlers require operator.approvals or operator.admin scopes. In affected releases, a scoped Gateway chat.send request delivered through an inherited external route could be evaluated as an external-channel command while still carrying the lower Gateway client scopes. This issue affects scoped Gateway clients. It does not apply to shared-secret bearer HTTP compatibility endpoints, which are documented as full operator surfaces under OpenClaw's trust model.

Sandboxed session spawn could expose the real workspace path to child prompts. In affected versions, a child session spawned from a sandboxed parent could forward the host workspace path into the child session prompt. This advisory is scoped to the named feature and configuration. It does not change OpenClaw's trusted-operator model: authenticated Gateway operators, installed plugins, and intentional local execution surfaces remain trusted unless a separate policy, approval, allowlist, sandbox, …

Same-host trusted-proxy deployments could accept local forged identity headers. In affected versions, a local same-host caller that can reach the proxy-facing Gateway port could supply identity headers normally reserved for the trusted proxy. This advisory is scoped to the named feature and configuration. It does not change OpenClaw's trusted-operator model: authenticated Gateway operators, installed plugins, and intentional local execution surfaces remain trusted unless a separate policy, approval, allowlist, sandbox, or …

QQBot streaming command could mutate config without explicit allowFrom. In affected versions, a QQBot sender reaching the affected command could change configuration without requiring an explicit non-wildcard allowlist entry. This advisory is scoped to the named feature and configuration. It does not change OpenClaw's trusted-operator model: authenticated Gateway operators, installed plugins, and intentional local execution surfaces remain trusted unless a separate policy, approval, allowlist, sandbox, or auth boundary is crossed.

QQBot pre-dispatch slash commands could skip allowFrom checks. In affected versions, a QQBot sender able to invoke slash commands could dispatch the command before applying the configured allowFrom policy. This advisory is scoped to the named feature and configuration. It does not change OpenClaw's trusted-operator model: authenticated Gateway operators, installed plugins, and intentional local execution surfaces remain trusted unless a separate policy, approval, allowlist, sandbox, or auth boundary is crossed.

OpenClaw's QQBot channel can deliver native approval buttons for exec and plugin approvals. In affected releases, the button callback path resolved approvals without enforcing the configured QQBot approver identity. The text command approval path used the authorization check; the issue was specific to native QQBot approval buttons.

QQBot admin commands could skip DM-only and allowFrom policy. In affected versions, a QQBot sender able to trigger the exported command could route admin commands without the QQBot-specific DM-only and allowFrom checks. This advisory is scoped to the named feature and configuration. It does not change OpenClaw's trusted-operator model: authenticated Gateway operators, installed plugins, and intentional local execution surfaces remain trusted unless a separate policy, approval, allowlist, sandbox, or auth …

PowerShell encoded-command aliases could miss exec allowlist checks. In affected versions, a command request using abbreviated encoded-command flags could use an alias form not recognized by the allowlist parser. This advisory is scoped to the named feature and configuration. It does not change OpenClaw's trusted-operator model: authenticated Gateway operators, installed plugins, and intentional local execution surfaces remain trusted unless a separate policy, approval, allowlist, sandbox, or auth boundary is crossed.

OpenClaw nodes send lifecycle events back to the gateway. In affected releases, a paired node could send an exec lifecycle event that was accepted without enough provenance tying it to an authorized system.run request. This issue affects the node event boundary. It does not allow an unauthenticated caller to reach the gateway; the attacker must already control a paired node connection.

The bundled device-pair plugin exposed /pair on normal chat command surfaces. In affected releases, authorized non-owner chat senders could issue device-pairing bootstrap codes without having owner, admin, or pairing scope. This issue does not affect unauthenticated users. The caller must already be allowed to send commands to the agent through a configured chat channel.

Node pairing reconnection could confuse approval scope state. In affected versions, a paired or reconnecting node session could mutate pairing state in a way that changed the approval scope decision. This advisory is scoped to the named feature and configuration. It does not change OpenClaw's trusted-operator model: authenticated Gateway operators, installed plugins, and intentional local execution surfaces remain trusted unless a separate policy, approval, allowlist, sandbox, or auth boundary is …

Native command authorization could skip owner-command enforcement. In affected versions, a sender able to trigger native command handling could authorize a native command without enforcing the configured owner-only command policy. This advisory is scoped to the named feature and configuration. It does not change OpenClaw's trusted-operator model: authenticated Gateway operators, installed plugins, and intentional local execution surfaces remain trusted unless a separate policy, approval, allowlist, sandbox, or auth boundary is …

message.action forwarding could send Gateway credentials to model-supplied loopback URLs. In affected versions, model-controlled action metadata that selects a loopback Gateway URL could forward the action payload with Gateway credentials to the supplied loopback URL. This advisory is scoped to the named feature and configuration. It does not change OpenClaw's trusted-operator model: authenticated Gateway operators, installed plugins, and intentional local execution surfaces remain trusted unless a separate policy, approval, allowlist, …

Message read actions could skip channel allowlist checks. In affected versions, a lower-trust caller with access to the affected message read action could request messages without the same channel allowlist check used by normal delivery. This advisory is scoped to the named feature and configuration. It does not change OpenClaw's trusted-operator model: authenticated Gateway operators, installed plugins, and intentional local execution surfaces remain trusted unless a separate policy, approval, allowlist, …

memory-wiki ingest could read local files with operator.write scope. In affected versions, a Gateway caller with operator.write access to the plugin tool could read arbitrary local file paths instead of staying within the intended ingest sources. This advisory is scoped to the named feature and configuration. It does not change OpenClaw's trusted-operator model: authenticated Gateway operators, installed plugins, and intentional local execution surfaces remain trusted unless a separate policy, approval, …

MCP loopback could skip owner-only tool policy for non-owner callers. In affected versions, a non-owner caller reaching the affected loopback path could skip owner-only tool policy and before-tool-call hooks. This advisory is scoped to the named feature and configuration. It does not change OpenClaw's trusted-operator model: authenticated Gateway operators, installed plugins, and intentional local execution surfaces remain trusted unless a separate policy, approval, allowlist, sandbox, or auth boundary is crossed.

Mattermost slash token revocation could lag until monitor refresh. In affected versions, a caller with an old Mattermost slash token during the refresh window could continue accepting the old token until the monitor refreshed. This advisory is scoped to the named feature and configuration. It does not change OpenClaw's trusted-operator model: authenticated Gateway operators, installed plugins, and intentional local execution surfaces remain trusted unless a separate policy, approval, allowlist, sandbox, …

Mattermost handlers could fall open when channel type was missing. In affected versions, a Mattermost event missing channel type metadata could continue without applying the intended DM policy decision. This advisory is scoped to the named feature and configuration. It does not change OpenClaw's trusted-operator model: authenticated Gateway operators, installed plugins, and intentional local execution surfaces remain trusted unless a separate policy, approval, allowlist, sandbox, or auth boundary is crossed.

Matrix allowFrom could bind to mutable display names. In affected versions, a Matrix account able to change display name metadata could match a policy entry through mutable display metadata. This advisory is scoped to the named feature and configuration. It does not change OpenClaw's trusted-operator model: authenticated Gateway operators, installed plugins, and intentional local execution surfaces remain trusted unless a separate policy, approval, allowlist, sandbox, or auth boundary is crossed.

OpenClaw hook ingress can start automated agent runs using a configured hook token. In affected releases, a hook-triggered run could select a bundled CLI backend that received owner-scoped MCP loopback authority instead of a scope appropriate for hook ingress. This issue affects the boundary between hook-token automation and owner-only MCP tools. It does not affect deployments with hooks disabled.

Feishu dynamic-agent bindings could miss configWrites enforcement. In affected versions, a Feishu sender using dynamic-agent binding behavior could create or update bindings without honoring the configured config-write control. This advisory is scoped to the named feature and configuration. It does not change OpenClaw's trusted-operator model: authenticated Gateway operators, installed plugins, and intentional local execution surfaces remain trusted unless a separate policy, approval, allowlist, sandbox, or auth boundary is crossed.

Fake package roots could influence memory-core artifact loading. In affected versions, a local package root resolution path influenced by workspace state could select a package root that was not the intended bundled artifact root. This advisory is scoped to the named feature and configuration. It does not change OpenClaw's trusted-operator model: authenticated Gateway operators, installed plugins, and intentional local execution surfaces remain trusted unless a separate policy, approval, allowlist, sandbox, …

OpenClaw exec approvals could show a shortened command in the approval UI while keeping the full original command for execution. For very long commands, an approver could see and approve a benign-looking prefix while a hidden suffix remained part of the command that would run after approval. This issue affects the approval display and binding for oversized exec commands. It does not make exec available to unauthenticated users, and it …

Embedded runner policy could be confused by provider aliases. In affected versions, a request using provider aliases could compare policy against an alias instead of the canonical provider identity. This advisory is scoped to the named feature and configuration. It does not change OpenClaw's trusted-operator model: authenticated Gateway operators, installed plugins, and intentional local execution surfaces remain trusted unless a separate policy, approval, allowlist, sandbox, or auth boundary is crossed.

In affected LAN/shared-token Control UI deployments, a caller could spoof locality information used during Control UI pairing and obtain a durable admin-capable device token. This issue is limited to deployments where the caller already has the network/authentication foothold needed to reach the Control UI pairing path. It is not an unauthenticated internet exposure issue.

Combined POSIX shell options could confuse exec revalidation. In affected versions, a command request using combined shell flags could parse approval-time and execution-time shell options differently. This advisory is scoped to the named feature and configuration. It does not change OpenClaw's trusted-operator model: authenticated Gateway operators, installed plugins, and intentional local execution surfaces remain trusted unless a separate policy, approval, allowlist, sandbox, or auth boundary is crossed.

Bundle MCP loopback could miss its exec denylist on session spawn. In affected versions, a caller that can reach the affected bundled MCP session-spawn path could bypass the denylist that was intended for that loopback MCP entry point. This advisory is scoped to the named feature and configuration. It does not change OpenClaw's trusted-operator model: authenticated Gateway operators, installed plugins, and intentional local execution surfaces remain trusted unless a separate …

Browser debug/export routes could reuse already-open blocked tabs. In affected versions, a caller that can reference an already-open browser tab could reuse blocked private-network tabs without reapplying the expected SSRF policy. This advisory is scoped to the named feature and configuration. It does not change OpenClaw's trusted-operator model: authenticated Gateway operators, installed plugins, and intentional local execution surfaces remain trusted unless a separate policy, approval, allowlist, sandbox, or auth boundary …

Stored XSS through wikitext can be performed by inserting malicious HTML into the overlays parameter of the display_map parser function when using the leaflet service.

All HTTP routes under /api/documents/* in mcp-memory-service are served without any authentication dependency, even when the server is configured with an API key (MCP_API_KEY) or OAuth. An unauthenticated remote attacker can upload arbitrary content into the memory store (write), retrieve stored document content (read), and permanently delete memories belonging to authenticated users (delete) — all without supplying any credentials. The /api/memories counterpart correctly enforces authentication, making this an inconsistent and …

A path traversal vulnerability exists in the campaign import feature of Mautic 7. When extracting uploaded ZIP files during campaign imports, a flaw in the validation logic allows file paths to escape the intended temporary directories.

A stored Cross-Site Scripting (XSS) vulnerability exists in the Projects component of Mautic 7. When displaying project tags and popovers on administrative detail views (such as campaigns, emails, or forms), user-supplied project names are rendered without proper sanitization. An authenticated user with permissions to create or edit projects can exploit this to inject malicious script payloads.

A stored Cross-Site Scripting (XSS) vulnerability exists in the project selector component of Mautic 7. When rendering selection menus for associating projects with system entities, the application fails to sanitize project names returned via AJAX before injecting them into the DOM as option fields. An authenticated user with permissions to create projects can exploit this to store a malicious script payload in the project's name.

An SQL injection vulnerability exists in Mautic's API contact filtering mechanism. Due to insufficient recursive sanitization of nested query parameters, an authenticated API user can bypass input filtering and inject arbitrary SQL commands.

A Server-Side Template Injection (SSTI) vulnerability exists in Mautic's theme engine. The platform renders uploaded Twig templates without a sandbox or strict function restrictions. Authenticated users with permissions to create or upload themes can abuse this to execute arbitrary code.

An authorization bypass vulnerability exists in the Mautic 7 API v2 endpoints (utilizing API Platform). Under certain conditions, roles configured with owner-scope restrictions (such as viewown or editown) are not properly enforced. This allows low-privilege authenticated API users to bypass ownership-logic controls and access or modify resources belonging to other users.

A Server-Side Request Forgery (SSRF) vulnerability exists in the Mautic Focus component (MauticFocusBundle). Under certain conditions, insufficiency in validating user-supplied URLs allows authenticated users to trigger outbound HTTP requests from the hosting server.

In the Debian.sudoers file, apt-get is allowed for the nagios user. The full command including the arguments are not enforced and can therefore be choosen arbitrarily. This allows to easily get a root shell as the nagios user:

An unauthenticated path traversal in the LaunchServer HTTP file server (FileServerHandler) lets any remote actor read any file readable by the LaunchServer process (e.g. ../../../../etc/passwd). This is a generic arbitrary-file-read primitive, so the fix must address the traversal itself, not any specific file. The readable files include the server's own secrets, which turns this from information disclosure into full compromise: the ECDSA private key that signs access JWTs (.keys/ecdsa_id), the …

SQLChatAgent in langroid ships a _validate_query defense-in-depth layer whose _DANGEROUS_SQL_PATTERNS regex blocklist enumerates dangerous SQL primitives by specific function name. The list misses the canonical PostgreSQL filesystem-disclosure family pg_read_file(), pg_stat_file(), pg_ls_logdir(), pg_ls_waldir(), pg_current_logfile() (and similar SELECT-shaped functions in the same family). It also leaves SQL Server OPENDATASOURCE and SQLite ATTACH '<file>' AS x (DATABASE keyword omitted) unblocked. An attacker able to shape the LLM's generated SQL (directly via prompt input …

Langroid's ReadFileTool and WriteFileTool appear to treat curr_dir as the intended working-directory boundary for file operations. However, the tools only change the process working directory to curr_dir and then operate on the user-supplied file_path without resolving and enforcing that the final path remains inside curr_dir. As a result, a tool caller can supply path traversal sequences such as ../secret.txt to read files outside the configured current directory, or ../written_by_tool.txt to …

Kiwi TCMS provides the /init-db/ page as part of its setup mechanism for administrators who prefer a browser instead of the command line. In previous versions of Kiwi TCMS this page still renders and responds to requests even after first use.

Kimai 2.56.0 contains an authenticated improper authorization / IDOR vulnerability in the favorite timesheet add and remove endpoints. A low-privileged user who knows another user's timesheet.id can add that record to, or remove it from, the victim's favorite/recent bookmark list. This allows cross-user manipulation of per-user favorite state without administrative privileges.

Keycloak's SAML broker endpoint does not properly validate encrypted assertions when the overall SAML response is not signed. An attacker with a valid signed SAML assertion can exploit this by crafting a malicious SAML response, injecting an encrypted assertion for an arbitrary principal, leading to unauthorized access and potential information disclosure.

The Kerberos Hub upload path sends the agent's Hub credentials in the custom X-Kerberos-Hub-PrivateKey and X-Kerberos-Hub-PublicKey request headers to the operator-configured Hub URL (config.HubURI). The HTTP client used (&http.Client{} in UploadKerberosHub) is constructed without a CheckRedirect policy, so it follows HTTP redirects automatically. Go's net/http strips only sensitive headers (Authorization, Cookie, WWW-Authenticate) on a cross-host redirect; it does not strip custom headers such as X-Kerberos-Hub-PrivateKey. As a result, if the …

Logic bug in decode_simple_table_slow may cause integer arithmetic overflow when decoding Modular image with certain kind of MA tree, which may panic with overflow-checks enabled.

jxl-oxide exposes a public safe API that can construct an undersized FrameBuffer due to unchecked usize multiplication, which immediately trigger panic while initializing the buffer in normal decoding path. Additionally, calling the safe grouped buffer accessors afterward can create invalid oversized slices from a much smaller allocation, causing undefined behavior; however normal decoding path never reaches UB, because these methods are never used within jxl-oxide.

On 32-bit platforms, decoding a crafted image may lead to out-of-bounds writes due to integer overflow in length calculation.

In JSONata <v2.2.0, it is possible to craft non-matching inputs to the $toMillis function that cause superlinear backtracking in the ISO-8601 validation regex. This may lead to denial of service in applications that evaluate user-provided JSONata expressions.

joserfc.jwt.decode accepts attacker-forged HMAC-signed tokens when the caller-supplied verification key is the empty string or None. HMACAlgorithm.sign and HMACAlgorithm.verify in src/joserfc/_rfc7518/jws_algs.py:62-70 feed whatever OctKey.get_op_key(…) produced into hmac.new(…), and OctKey.import_key only emits a SecurityWarning when the raw key is shorter than 14 bytes without rejecting zero-length input. Any application whose JWT secret is sourced from an unset environment variable, an unset Redis / DB row, a key finder fallback that returns …

Authorization for scoped (agent) MCP callers is enforced inline, per tool, and is applied inconsistently — several mutating tools silently omit the ancestry/workspace check that their siblings perform. Because the MCP server authenticates all outbound gRPC with the full server API key and the backend gRPC handlers perform no caller-based authorization, the MCP tool layer is the sole authorization boundary. A malicious or prompt-injected scoped agent can therefore perform cross-task …

Authorization for scoped (agent) MCP callers is enforced inline, per tool, and is applied inconsistently — several mutating tools silently omit the ancestry/workspace check that their siblings perform. Because the MCP server authenticates all outbound gRPC with the full server API key and the backend gRPC handlers perform no caller-based authorization, the MCP tool layer is the sole authorization boundary. A malicious or prompt-injected scoped agent can therefore perform cross-task …

Authorization for scoped (agent) MCP callers is enforced inline, per tool, and is applied inconsistently — several mutating tools silently omit the ancestry/workspace check that their siblings perform. Because the MCP server authenticates all outbound gRPC with the full server API key and the backend gRPC handlers perform no caller-based authorization, the MCP tool layer is the sole authorization boundary. A malicious or prompt-injected scoped agent can therefore perform cross-task …

The default git executor used for all worktree operations spawns git through a shell, and the untrusted task branch name flows into the command unsanitized. A caller able to reach the PowerLine SpawnSession RPC (a malicious or compromised agent acting through the orchestration layer, or any client able to spawn a task) can achieve arbitrary command execution as the PowerLine user on every provisioned environment (SSH host, Docker container, or …

The default git executor used for all worktree operations spawns git through a shell, and the untrusted task branch name flows into the command unsanitized. A caller able to reach the PowerLine SpawnSession RPC (a malicious or compromised agent acting through the orchestration layer, or any client able to spawn a task) can achieve arbitrary command execution as the PowerLine user on every provisioned environment (SSH host, Docker container, or …

The TIFF decoder does not place a limit on the size of PackBits-compressed data. A maliciously-crafted image can exploit this to cause a small image (both in terms of pixel width/height and encoded size) to make the decoder decode large amounts of compressed data.

The BalancerForward proxy helper in GoFiber uses Header.Add() instead of Header.Set() when injecting the X-Real-IP header. This appends the real client IP as a second header value rather than replacing any attacker-supplied value. Upstream servers that read the first X-Real-IP header (nginx, Express, most HTTP servers) use the attacker's spoofed IP for logging, rate limiting, and access control.

The BalancerForward proxy helper in GoFiber uses Header.Add() instead of Header.Set() when injecting the X-Real-IP header. This appends the real client IP as a second header value rather than replacing any attacker-supplied value. Upstream servers that read the first X-Real-IP header (nginx, Express, most HTTP servers) use the attacker's spoofed IP for logging, rate limiting, and access control.

The default Authorizer function in GoFiber's BasicAuth middleware uses short-circuit evaluation that skips password hash comparison for non-existent usernames. With bcrypt-hashed passwords (the primary use case), the timing difference between a valid and invalid username is approximately 1,000,000:1 (~100ms vs ~100ns), enabling reliable remote username enumeration.

An authenticated customer can read other customers' allowed sender aliases from Froxlor's sender-delete confirmation page when mail.enable_allow_sender is enabled. customer_email.php loads allowed_sender by global auto-increment senderid alone, so a customer can enumerate foreign sender alias IDs and make Froxlor disclose those values in the confirmation dialog for the attacker's own mailbox.

The Mysqls.add API command (lib/Froxlor/Api/Commands/Mysqls.php) accepts a customer-controlled mysql_server parameter and only validates that the value is numeric and that the server index exists in userdata.inc.php. It never checks the value against the calling customer's allowed_mysqlserver allowlist. A customer can therefore create a database, plus a MySQL user with a password they choose, on any MySQL server the operator has configured — including servers that were explicitly excluded from that …

fast-mcp-telegram validates HTTP Bearer tokens by joining the raw token string into a session-file path. The verifier rejects the exact reserved token telegram, but it does not reject path separators or normalize the path before checking whether the session file exists. A remote HTTP client can therefore authenticate as the default legacy session with a token such as ../fast-mcp-telegram/telegram when the documented default session file ~/.config/fast-mcp-telegram/telegram.session exists. This bypasses the …

A path traversal vulnerability exists in the Zmodem and Trzsz file download handlers in electerm. When receiving files via Zmodem or Trzsz protocols, electerm uses the remote-supplied filename directly in path.join() with the user-selected download directory without sanitization. A malicious SSH server or remote shell process can send a specially crafted filename such as ../escaped.txt to escape the user-selected download directory and write files to arbitrary locations on the user's …

A command injection vulnerability exists in electerm's file system operations (rmrf, mv, cp) in src/app/lib/fs.js. These functions construct shell commands by interpolating file paths directly into command strings without escaping shell metacharacters. Vulnerable functions: rmrf() - Uses rm -rf "${path}" (double quotes, vulnerable to " injection) mv() - Uses mv '${from}' '${to}' (single quotes, vulnerable to ' injection) cp() - Uses cp -r "${from}" "${to}" (double quotes, vulnerable to " …

dulwich.porcelain.submodule_update, and by extension porcelain.clone(…, recurse_submodules=True), materializes attacker-controlled submodule paths from a crafted upstream repository without path validation. A malicious .gitmodules plus a matching tree gitlink whose path is .git/hooks (or any other directory inside the parent repository's .git directory) causes the attacker's submodule tree contents to be written directly into the victim's .git/hooks/ directory, preserving executable mode bits. The dropped executables are then run by any subsequent git or …

The Dragonfly Manager exposes GET /api/v1/oauth and GET /api/v1/oauth/:id to unauthenticated clients. The response body deserializes the entire manager/models.Oauth struct, which includes the client_secret field. Any network-reachable attacker can read the OAuth client secrets configured for github or google providers, defeating the confidentiality guarantee of those secrets and enabling subsequent abuse against the connected identity providers.

There is a mass-assignment flaw in the bulk-duplicate element action. Alice, holding only the permission to duplicate an entry she owns, submits an arbitrary id through the newAttributes request parameter. The duplication routine overrides its own id = null reset with that value and writes Alice’s attributes into Bob’s existing entry row.

We have identified an authorization issue in Craft CMS AssetsController::actionReplaceFile that can delete a source asset without source delete permission by supplying both assetId and sourceAssetId.

AssetsController::actionDeleteFolder() only requires the deleteAssets:<volume-uid> permission for the target folder. It never enforces deletePeerAssets:<volume-uid>, even though Assets::deleteFoldersByIds() cascades deletion to every descendant folder and every asset inside, regardless of who uploaded them. A low-privilege user who has been granted folder-management rights on a shared volume can therefore destroy assets uploaded by other users (peer assets), bypassing the per-asset peer-permission check that the sibling actionDeleteAsset endpoint correctly applies. This is the …

EntriesController::actionSaveEntry() performs entry-edit permission checks before request-controlled author changes are applied to the model. The subsequent author mutation path accepts attacker-supplied authors / author parameters and allows the change when the current user is one of the old authors. Because the controller does not re-run authorization after mutating the author list, a low-privileged user can reassign an entry’s authorship to another user without holding the dedicated peer-author-change permission.

The EntriesController::actionMoveToSection() endpoint checks only whether the current user can view the destination section, but it does not require permission to save entries into that section. A low-privileged authenticated control-panel user who can move an entry out of its current section can therefore move that entry into a different section where they have read access but no write access.

We have identified an authorization issue in Craft CMS where a forced folder move can delete a conflicting destination folder without destination delete permission.

When an HTTPProxy is configured with incompatible combination of both .spec.virtualhost.tls.enableFallbackCertificate: true and .spec.virtualhost.jwtProviders, Contour does not reject the configuration. Consequently, requests from clients that do not send TLS SNI or send an unrecognized SNI (one that does not match any HTTPProxy FQDN) bypass configured JWT verification and are proxied to upstream services without a valid token. To list all HTTPProxies with this invalid configuration, run kubectl get httpproxies -A …

The dotfiles registry module passed unsanitized user input to shell commands, allowing arbitrary code execution inside a provisioned workspace. Any user who supplied a crafted dotfiles_uri value (for example, one containing shell command substitution such as $(…)) could achieve command execution in their own workspace. The Create Workspace page's mode=auto deep links amplified this into a one-click attack: an attacker could craft a URL that prefilled param.dotfiles_uri and silently provisioned …

The dotfiles registry module passed unsanitized user input to shell commands, allowing arbitrary code execution inside a provisioned workspace. Any user who supplied a crafted dotfiles_uri value (for example, one containing shell command substitution such as $(…)) could achieve command execution in their own workspace. The Create Workspace page's mode=auto deep links amplified this into a one-click attack: an attacker could craft a URL that prefilled param.dotfiles_uri and silently provisioned …

The aarch64 implementations of Cmov and CmovEq seem to assume that the high bits when loading a value of size smaller than a register into a register are zero-extended. However, this is not the case and these bits are unspecified. This can result in a left.cmovz(&right, condition) not moving right into left, even if condition == 0.

Algernon selects its file handler from filepath.Ext() (engine/handlers.go:134), which does not treat the NTFS-equivalent names x.lua::$DATA, x.lua., or x.lua as .lua. On Windows, an unauthenticated client appends one of these suffixes to any server-side script on a public path and receives its raw source instead of executed output, leaking embedded secrets such as database credentials and the SetCookieSecret value. Linux and macOS hosts are unaffected.

9router uses a publicly known hardcoded string "9router-default-secret-change-me" as the fallback of JWT secret for all Dashboard session JWTs when the JWT_SECRET environment variable is not set. Because this secret is committed in the public repository and unchanged across all releases, any unauthenticated remote attacker can forge a valid auth_token cookie and gain full access to dashboard and api (If JWT_SECRET is not set on server) . This vulnerable affected …

POST /api/tunnel/tailscale-install accepts a JSON body with a sudoPassword field and pipes it, followed by the body of https://tailscale.com/install.sh, into a child process spawned as sudo -S sh. The route is not present in the dashboard middleware matcher in src/proxy.js, so the request reaches the handler without invoking dashboardGuard.proxy(). In deployments where the Node process runs as root (Docker images derived from node:* without a USER directive, npm i -g …

The fix for CVE-2026-46339 (unauthenticated RCE via unprotected MCP plugin routes) introduced a local-only access gate in src/dashboardGuard.js that restricts spawn-capable routes (/api/mcp/, /api/tunnel/, /api/cli-tools/*) to loopback requests. The gate determines "local" by inspecting the Host and Origin HTTP headers rather than the TCP source address. When 9router is deployed behind a reverse proxy, tunnel (Cloudflare Tunnel, Tailscale — both natively supported), or is subject to DNS rebinding, these headers …

UForm and UAuthForm render a server-side <form> element with no method and no action attribute, relying on a hydrated @submit.prevent handler to intercept submission. If a user submits the form before Vue hydration has attached the handler (autofill plus Enter on a slow network, JS bundle blocked by CSP or CDN failure, etc.), the browser performs the native default: a GET to the current URL with every named field, including …

A CPU exhaustion vulnerability exists in Conform's parseSubmission future API when parsing FormData or URLSearchParams submissions with many unique field names. The parser previously looked up values by field name, which could require repeated scans of the submitted entries and cause excessive synchronous CPU work if an attacker supplies a crafted submission. [!NOTE] The patched version fixes this by iterating submitted entries directly instead of repeatedly looking up values by …

Finding Location: core/src/shared/secure-fetch.ts:42-45 When new URL() throws a parse error, the assertSecureUrl function returned without throwing, silently allowing the request to proceed without HTTPS validation. Status Fixed in v0.2.136 — The catch block now throws an error instead of silently returning.

Finding Location: core/src/core/scheduler.ts:23, core/src/hooks/dispatcher.ts:100, core/src/client/graphql.ts:71 Several console.warn calls are not gated behind DEV and will fire in production builds, potentially exposing internal framework state such as queue sizes, component names, and query fragments to users viewing the browser console. Status Open — These warnings serve as development-time diagnostics. They do not expose credentials or PII, but may reveal internal architecture details. Recommendation Gate all development-time console.warn and console.error calls behind …

Finding Location: core/src/shared/secure-fetch.ts assertSecureUrl validated only the initial request URL. The fetch() API follows redirects by default (up to 20 hops). A request to a valid https:// URL could redirect to http://internal-service/ or other unvalidated destinations. Status Fixed in v0.2.136 — secureFetch now defaults to redirect: 'error' which rejects any redirect. Callers can override with { redirect: 'follow' } if they trust the target.

Finding Location: core/src/shared/secure-fetch.ts:52-54 The localhost exception allowed localhost and 127.0.0.1 but did not cover 0.0.0.0, [::1] (IPv6 localhost), or the full 127.0.0.0/8 loopback range. Status Fixed in v0.2.136 — Localhost detection now covers localhost, 127.0.0.1, [::1], 0.0.0.0, and the full 127.x.x.x range.

Finding Location: core/src/client/graphql.ts:66-80 The gql template tag function warned about interpolated values containing GraphQL metacharacters ({}():) but still concatenated them into the query string, enabling potential GraphQL injection. Status Fixed in v0.2.136 — The gql function now throws an error when metacharacters are detected in interpolated values, forcing developers to use the variables parameter.

Finding Location: core/src/server/render-to-string.ts:307-311 CSS value sanitization stripped expression( and url(javascript: using simple regex, but could be bypassed with CSS unicode escapes (\65xpression(), null bytes, or CSS comments (exp/**/ression(). Mitigating Factor: These CSS injection vectors only work in legacy browsers (IE6-IE10). SpecifyJS targets modern browsers. Status Fixed in v0.2.136 — CSS sanitization now normalizes unicode escapes and strips CSS comments before pattern matching. Also checks for behavior:, -moz-binding, and -o-link patterns.

Finding Location: core/src/shared/secure-fetch.ts:33-35 data: URIs were allowed without any restriction. While data: URIs don't make network requests, they can be used for memory exhaustion via very large data URIs. Status Fixed in v0.2.136 — data: URIs are now limited to 1MB. URIs exceeding this limit throw an error.

The matchFileSnapshot function in src/assertions/snapshots.ts accepted a filePath parameter with zero validation. When snapshot update mode was active (UPDATE_SNAPSHOTS=1 or setUpdateMode('all')), an attacker who controls test input could write arbitrary content to any filesystem path the process has write access to, including creating intermediate directories.

The HTMLInputElement.checkValidity() method constructed a RegExp directly from the user-controlled pattern property without any sanitization or timeout protection. This allowed an attacker to inject a regex with catastrophic backtracking, freezing the event loop.

The wetty client decodes a base64 filename from the file-download escape sequence and interpolates it raw into a Toastify HTML string (escapeMarkup: false). Any output the victim renders - a cat'd file, a tailed log, an SSH MOTD, a curl response - that contains \x1b[5i…:…\x1b[4i runs script in the wetty origin and types attacker-chosen keystrokes into the victim's SSH session.

The per-template filter, tag and function allow-list check is compiled into the checkSecurity() method of each Template subclass and was invoked once from the constructor, gated by SandboxExtension::isSandboxed($source). Template instances are then cached on the Environment in $loadedTemplates, so the verdict computed at construction time was sticky for the rest of the process. Any later change of sandbox state on the same Environment left that cached verdict in place: toggling …

An anonymous caller could create new namespaces and databases on a running SurrealDB instance without holding DEFINE NAMESPACE or DEFINE DATABASE permission. USE NS <name> and USE DB <name> automatically create the target when it does not exist. The three places USE is handled — the RPC use method, Datastore::process_use, and the SurrealQL executor — did not check whether the caller was allowed to create the resource. Under default capabilities …

A vulnerability was discovered where the user-supplied WHERE clause in a SELECT statement is evaluated against the full record data before PERMISSIONS FOR SELECT WHERE determines whether the principal is authorised to access that record. A side-effecting expression in the WHERE clause can exfiltrate record contents before the permission check runs. The same ordering bug affects the SET, MERGE, CONTENT and PATCH clauses of update-variant statements (UPDATE, UPSERT-update, INSERT ON …

SurrealDB offers http::* functions that can access external network endpoints, with the –allow-net and –deny-net capabilities used to restrict the set of network targets that can be reached. An authenticated user of SurrealDB can bypass a port-scoped –deny-net <host>:<port> rule by chaining an HTTP redirect: the initial request goes to an –allow-net-permitted hostname, the response's 3xx Location header points at the denied host:port, and the redirect is followed even though …

A LIVE SELECT subscription records the user's auth state ($auth, $token, $session, $access) when it is registered, and the server uses that recorded state to evaluate the table- and row-level PERMISSIONS clauses for every subsequent notification. The recorded state is never refreshed. When something changes the user's effective auth state — the originating session is invalidated, the session's TTL expires, or the user signs in, signs up, or authenticates as …

The HTTP /rpc endpoint has a time-of-check/time-of-use (TOCTOU) race condition on internal session state. When authenticated and unauthenticated requests are processed concurrently, the unauthenticated request can inherit the authenticated user's session and privileges. The /rpc endpoint is the primary interface used by all official SurrealDB SDKs. The HTTP /rpc handler does not bind each incoming request to an isolated session context. Instead, concurrent requests share mutable authentication state. When an …

The HTTP /rpc sessions method returned every attached session UUID without authentication, and the /rpc handler accepted an arbitrary session field with no ownership check. An anonymous caller could enumerate UUIDs and impersonate any authenticated session. "Attached" means sessions registered via {"method":"attach"} — the only writer to the HTTP session map. Ordinary stateless /rpc requests use ephemeral per-request sessions that are filtered from sessions() and destroyed at end-of-request, so they …

An authenticated record or scope user could read records on any table reachable through a graph edge or REFERENCES TO back-reference, regardless of that table's PERMISSIONS FOR select clause. Traversing SELECT * FROM source->edge->target returned full documents from target even when target was defined as PERMISSIONS FOR select NONE. The same bypass extended through multi-hop chains, so any table reachable by a sequence of edges from a readable starting point …

A record user could learn the value of a hidden field by counting how many records match a guess. When DEFINE FIELD … PERMISSIONS FOR select WHERE … hides a field's contents from a caller, and that field is indexed, running SELECT count() FROM t WHERE hidden_field = "guess" GROUP ALL returned a count greater than zero whenever a record actually had that value — even though the caller was …

When a user configures ALGORITHM ES512 for any JWT access method (DEFINE ACCESS … TYPE JWT ALGORITHM ES512), SurrealDB silently substitutes ES384 at all four internal algorithm conversion points. This occurs because the underlying jsonwebtoken crate (v10.x) does not include an ES512 algorithm variant, so the mapping defaults to ES384 without raising an error, warning, or log message. Users who provide the correct P-521 key type for ES512 will experience …

In SurrealDB, records can be connected as a graph: a RELATE statement creates an edge record between two node records. If either endpoint node is deleted, SurrealDB automatically removes the edge row to keep the graph consistent. A user with permission to delete a node could also delete the edges connected to that node, even when the edge table's PERMISSIONS FOR delete clause should have stopped them. The automatic edge …

A LIVE query whose WHERE clause evaluates to an error caused the source data modifier (the user creating, updating, or deleting a record on the watched table) to fail instead. Calling any arbitrary SurrealQL function with a typed parameter and passing a value of the wrong type — for example LIVE SELECT * FROM t WHERE string::trim(deny) — triggered an evaluation error inside the LIVE notification path. That error then …

The KILL statement is used to terminate LIVE SELECT subscriptions that capture real-time changes to data within a table. The KILL statement implementation in core/src/expr/statements/kill.rs verifies that the requesting user has database-level access, but does not verify that the requesting user is the owner of the live query being terminated. After passing the valid_for_db() check, the KILL statement resolves the live query UUID, looks up the corresponding live query entry, …

A record user could read records the table's SELECT permission expression should have hidden, when that expression referenced $value, $before, $after, or $event. Binding a chosen value to that name before registering a LIVE SELECT caused notifications to evaluate the permission against the attacker's input instead of the real document.

A record user with UPDATE access could read field values that field-level SELECT permissions hid from them. Arithmetic operators and extend embedded the raw operand into their error messages, and UPDATE permission checks evaluate against the unreduced document — so triggering such an error against a hidden field returned its value in the resulting error.

RELATE creates an edge record between two existing records, and SurrealDB enforces the CREATE permission on the edge table for this operation. When the statement included a SET id = edge:existing clause, however, the new edge's id ended up pointing at an record that was already in storage. Rather than failing because the target already existed — which is what a create operation should do — the storage layer silently …

An anonymous caller could degrade /sql availability by streaming WebSocket frames many times larger than the operator-configured per-connection limit. The /sql upgrade handler accepted anonymous connections and did not propagate SURREAL_WEBSOCKET_MAX_MESSAGE_SIZE to the WebSocket protocol layer — incoming bytes accumulated in the per-connection read buffer before check_anon could reject the query, so the memory cost was incurred regardless of whether the caller could ever execute SurrealQL. The same upgrade path …

The SurrealDB type/kind parser did not enforce the configured recursion depth limit when parsing nested type annotations. The expression parser already enforced the limit for analogous constructs; the kind parser omitted it. An authenticated attacker could send a query with deeply nested type annotations (e.g., array<option<array<option<…>>>>) and exhaust server memory, crashing the process.

A single unauthenticated WebSocket message to /rpc crashed the SurrealDB server. Sending use { db: "x" } without first selecting a namespace hit .expect("namespace should be set") in the use handler; because surrealdb-core is built with panic = 'abort', the panic terminated the process. use is callable before signin, and the per-method capability check passes by default for guest callers — so no credentials, token, or –allow-guests flag are required.

The SurrealDB value and JSON parser did not enforce the configured recursion depth limit when parsing nested {, [, or ( tokens. The expression parser already enforced the limit for these tokens; the value/JSON parser omitted it. An unauthenticated attacker could send a deeply nested JSON payload to the WebSocket /rpc endpoint and exhaust server memory, crashing the process.

SurrealDB lets callers modify records using JSON Patch operations via the UPDATE … PATCH statement (and SDK equivalents such as db.patch()). One of those operations is copy, which duplicates one field's value into another field of the same record. A PATCH with an empty from — for example, UPDATE thing:1 PATCH [{ op: 'copy', from: '', path: '/leak' }] — was treated as "copy the entire record" and duplicated every …

An authenticated user could bypass permission rules that gated access on parts of a record's id — most commonly tenant-isolation rules of the form PERMISSIONS FOR select WHERE id.tenant = $auth.id.tenant. The same defect also let UNIQUE constraints defined on parts of an id admit duplicate entries. When a query referenced part of a composite record id (id.tenant, id.uid, …), SurrealDB read the value from the record's editable body fields …

The documented certificateOIDs option in sigstore.verify() is accepted by the public API but discarded before verification, so required certificate extension OIDs are never checked.

sigstore-js derives a transparency-log timestamp from tlogEntries[].integratedTime and uses it to validate certificate validity windows and satisfy timestampThreshold. For bundle v0.2, a tlog entry can be inclusionProof-only (no signed inclusionPromise/set), and the inclusion proof path does not cryptographically bind integratedTime. As a result, an attacker who can supply an untrusted bundle can influence time-based verification decisions by choosing integratedTime.

Schema.org has a cross-site scripting (XSS) vulnerability via script break-out in toScript() output.

Repomix's MCP server exposes a normal file_system_read_file tool that reads absolute paths only after running the project's secret check. However, the attach_packed_output plus read_repomix_output flow can read arbitrary local .json, .txt, .md, or .xml files without the same safety check and without verifying that the file is actually a Repomix packed output. This is a medium-severity local MCP file-read boundary issue. The affected deployment is the documented repomix –mcp stdio …

The –remote-branch CLI option in repomix is vulnerable to argument injection. User-supplied input is passed directly to git fetch and git checkout subprocesses via child_process.execFileAsync without sanitization, – delimiters, or validation. An attacker can inject arbitrary git command-line options. By injecting the –upload-pack option and specifying an SSH (git@…) or local (file://) remote URL, an attacker achieves arbitrary command execution with the privileges of the user running repomix. This bypasses …

A critical command injection vulnerability has been identified in the Rancher Manager cluster import endpoint /v3/import/{token}_{clusterId}.yaml through unsanitized YAML parameters. This endpoint accepts an authImage query parameter that is rendered without sanitization into a generated Kubernetes manifest template. By including URL-encoded newlines in the parameter value, an attacker can break out of the image: field to inject arbitrary YAML keys and malicious configurations, such as commands to execute malicious containers. …

A vulnerability has been identified in Rancher Manager that allows users assigned the Project Owner role to modify Pod Security Admission (PSA) labels on namespaces within their projects. Under the default role configuration, an attacker with the following access pattern can exploit this issue: Cluster Access: The user is granted Cluster Member access. Project Ownership: The user creates or is assigned ownership of a project. Namespace Creation: The user creates …

A vulnerability has been identified within Rancher Manager in the GitHub App authentication provider. When evaluating permissions, the provider incorrectly expands user team memberships to include all teams within the associated GitHub organization, rather than restricting access to the specific teams to which the user actually belongs. Specifically, when a user authenticates via the GitHub App provider, Rancher's team membership evaluation logic incorrectly handles cached data. Instead of checking the …

A vulnerability in Fleet for Rancher Manager affects multi-tenancy environments where different tenants share the same downstream clusters (e.g., different privileged or untrusted teams inside the same organization). On unpatched versions, tenants could bypass restrictions to access any config map or secret across all namespaces on the downstream cluster. They can create cluster-wide resources using HelmOp or Bundle without authorization. Specifically, an attacker can exploit this vulnerability in the following …

A vulnerability has been identified in Fleet when the webhook endpoint is configured without a secret; an attacker can forge webhook requests. The attacker doesn't need to know the specific repository or path configured in the GitRepo resource to make Fleet process these requests. An attacker can exploit this vulnerability to cause the following impacts: Trigger continuous repository re-cloning, which increases network traffic and can deplete resources on the management …

A vulnerability has been identified in Fleet when the helmRepoURLRegex field isn't set on a GitRepo resource. Fleet's bundle reader forwards Helm authentication credentials (BasicAuth) to any URL specified in the helm.repo field of a fleet.yaml file. An attacker with git push access to a Fleet-monitored repository can exploit this behavior by specifying a malicious URL in helm.repo. This causes the Fleet controller to send the configured Helm repository credentials …

Pay::Webhooks::PaddleBillingController#valid_signature? (app/controllers/pay/webhooks/paddle_billing_controller.rb) verifies the Paddle Billing webhook signature by computing OpenSSL::HMAC.hexdigest(…) and comparing it to the attacker-supplied header value using Ruby's String#==. Ruby's == is non-constant-time — it returns as soon as the first byte mismatches — and exposes a per-byte timing side channel on the webhook signature verification path. The canonical mitigation is to use a constant-time primitive (OpenSSL.fixed_length_secure_compare / ActiveSupport::SecurityUtils.secure_compare).

oras-go's auth.Client follows the realm URL from a registry's WWW-Authenticate: Bearer challenge without validating its scheme or host. The realm field is server-controlled by design in the OCI/distribution spec — registries legitimately point token requests at a separate auth endpoint (e.g. Docker Hub's registry-1.docker.io -> auth.docker.io), so cross-host realms on public DNS names are not in themselves a vulnerability. Two specific patterns, however, are never legitimate under any registry trust …

oras-go's auth.Client follows the realm URL from a registry's WWW-Authenticate: Bearer challenge without validating its scheme or host. The realm field is server-controlled by design in the OCI/distribution spec — registries legitimately point token requests at a separate auth endpoint (e.g. Docker Hub's registry-1.docker.io -> auth.docker.io), so cross-host realms on public DNS names are not in themselves a vulnerability. Two specific patterns, however, are never legitimate under any registry trust …

The file content store in oras-go attempts to confine writes to workingDir when AllowPathTraversalOnWrite=false, but the guard is lexical and does not account for symlink traversal. If workingDir contains a symlink path component and an attacker-controlled blob title (via ocispec.AnnotationTitle) targets a path under that symlink, pushFile() can create a file outside workingDir.

oras-go follows a registry-controlled Location header during the monolithic blob upload flow and reuses the Authorization header from the initial POST request for the subsequent PUT request. If a malicious registry returns a cross-host Location, oras-go can send the caller's credentials to an attacker-controlled endpoint.

ORAS Go can forward registry credentials configured for one registry origin to a different HTTP origin during registry redirects. There are two related paths: A manifest or metadata request authenticates to the origin registry, then the origin returns a redirect to another host or port. The redirected request can carry the origin Authorization header to the redirect target. A blob upload POST authenticates to the origin registry, then the origin …

MCP SSE redirects could forward Authorization headers. In affected versions, a lower-trust caller or configured input path could execute or persist actions beyond the caller's intended authorization. This advisory is scoped to the named feature and configuration. It does not change OpenClaw's trusted-operator model: authenticated Gateway operators, installed plugins, and intentional local execution surfaces remain trusted unless a separate policy, approval, allowlist, sandbox, or auth boundary is crossed.

A memory-safety vulnerability in Open Babel's PQS parser caused an uninitialized pointer dereference when reading a crafted input file.

A memory-safety vulnerability in Open Babel's MSI parser caused an uninitialized pointer dereference when reading a crafted input file.

A memory-safety vulnerability in Open Babel's GRO parser caused an uninitialized pointer dereference when reading a crafted input file.

A memory-safety vulnerability in Open Babel's PQS parser allowed an out-of-bounds write when reading a crafted input file.

A memory-safety vulnerability in Open Babel's ORCA parser allowed an out-of-bounds write when reading a crafted input file.

A memory-safety vulnerability in Open Babel's ORCA parser allowed an out-of-bounds write when reading a crafted input file.

A memory-safety vulnerability in Open Babel's MSI parser allowed an out-of-bounds write into the translationVectors[] array when reading a crafted input file.

A memory-safety vulnerability in Open Babel's MOPAC output parser allowed an out-of-bounds write into the translationVectors[] array when reading the "FINAL POINT" block of a crafted input file.

A memory-safety vulnerability in Open Babel's MOPAC input parser allowed an out-of-bounds write into the translationVectors[] array when reading Tv (translation-vector) atoms from a crafted input file.

A memory-safety vulnerability in Open Babel's MOL2 parser allowed an out-of-bounds write when reading a crafted input file.

A memory-safety vulnerability in Open Babel's Gaussian output parser allowed an out-of-bounds write into the translationVectors[] array when reading a crafted input file.

A memory-safety vulnerability in Open Babel's Gaussian output parser allowed an out-of-bounds write when reading a crafted input file.

A memory-safety vulnerability in Open Babel's CSR parser allowed an out-of-bounds write when reading a crafted input file.

A memory-safety vulnerability in Open Babel's PQS parser caused an out-of-bounds (pre-buffer) read when reading a crafted input file.

A memory-safety vulnerability in Open Babel's ChemKin parser caused a NULL pointer dereference when reading a crafted input file.

A memory-safety vulnerability in Open Babel's CACAO parser caused a NULL pointer dereference when reading a crafted input file.

A memory-safety vulnerability in Open Babel's ChemKin parser caused a heap buffer overflow when reading a crafted input file.

A flaw in com.ongres.scram:scram-client allows an attacker capable of performing a TLS man-in-the-middle (MITM) attack to silently downgrade a connection from SCRAM-SHA-256-PLUS (with channel binding) to standard SCRAM-SHA-256 (without channel binding), bypassing strict client-side enforcement policies.

A flaw in com.ongres.scram:scram-client allows an attacker capable of performing a TLS man-in-the-middle (MITM) attack to silently downgrade a connection from SCRAM-SHA-256-PLUS (with channel binding) to standard SCRAM-SHA-256 (without channel binding), bypassing strict client-side enforcement policies.

The fix for GHSA-fpxj-m5q8-fphw (CVE-2026-45710, "Mailpit: Set a default 50MB p/m limit to prevent DoS via unlimited SMTP DATA and /api/v1/send body sizes") wrapped only POST /api/v1/send with http.MaxBytesReader. The four other Mailpit JSON-body API endpoints PUT /api/v1/messages (SetReadStatus), DELETE /api/v1/messages (DeleteMessages), PUT /api/v1/tags (SetMessageTags), and POST /api/v1/message/{id}/release (ReleaseMessage) still call json.NewDecoder(r.Body) directly with no body-size cap and remain reachable unauthenticated in the default docker run axllent/mailpit:latest deploy. An unauthenticated …

ArchiveUtils.untar(InputStream, Path) and ArchiveUtils.unzip(InputStream, Path) in land.oras:oras-java-sdk create symbolic-link entries from an archive without validating the symlink target. A malicious tar (or zip) shipped as an OCI layer blob can place a symlink under the extraction directory whose target points outside that directory, then ship a regular-file entry whose path traverses through that symlink. The subsequent Files.newOutputStream call follows the symlink and writes the file outside the caller's chosen target …

The LoginLink signature used for password reset URLs covers only the user's id — it does not include the password hash. After a user clicks a reset link and successfully changes their password, the same link remains valid for up to 2 additional uses within a 1-hour window. Anyone who intercepts or caches the original link can log in as the user even after the password has been changed.

A flaw was found in Keycloak's Fine-Grained Admin Permissions (FGAPv2) feature. An administrator with limited client management permissions can exploit this vulnerability to assign any realm role, including highly privileged roles, to a client's scope mapping. This bypasses intended security controls, allowing the injected role to be projected into a user's authentication token when they access the modified client. This could lead to unauthorized privilege escalation within the Keycloak realm.

When goshs is launched with WebDAV enabled (-w), the mode-restriction flags –read-only, –upload-only, and –no-delete are enforced only on the primary HTTP port. The WebDAV port is wired straight to golang.org/x/net/webdav.Handler with no equivalent guard, so an authenticated WebDAV client can PUT, DELETE, MKCOL, MOVE, and COPY despite the operator's stated intent.

ShareHandler reads the share token's DownloadLimit under RLock, releases the lock, serves the file, then re-acquires the lock to increment the counter. Concurrent requests all read the same Downloaded/DownloadLimit snapshot, all pass the check, and all are served — exceeding the operator's intended cap.

When Ghost is behind a shared caching layer that results in cached content being shared between different visitors (e.g., Fastly, Cloudflare, nginx proxy_cache, and others), an unauthenticated user could send an x-ghost-preview header that altered the rendered frontend response. In affected cache configurations, that response could be stored and served to subsequent visitors requesting the same page, allowing cache poisoning of request-specific preview output. When running Ghost's frontend and admin …

It is possible to craft a URL that causes GeoNetwork to reflect attacker-controlled content into an error page in a way that gets evaluated as a client-side template expression. Combined with known AngularJS sandbox-escape techniques, this can be used to execute arbitrary JavaScript in the victim's browser (reflected Cross-Site Scripting via client-side template injection).

GeoNetwork's Elasticsearch-backed search API is responsible for injecting access-control and visibility filters into every request before it reaches the underlying Elasticsearch index. Under certain request conditions, that filtering step does not run, allowing an unauthenticated user to retrieve indexed metadata records that should be restricted, including records limited to specific groups.

A vulnerability has been identified in Fleet's agent-side deployer, which did not filter security-sensitive keys from namespaceLabels in fleet.yaml (or BundleDeployment.spec.options.namespaceLabels) when applying them to the target namespace. An attacker with git push access to a Fleet-monitored repository could overwrite Pod Security Standards (PSS) enforcement labels on a target namespace. This allows the attacker to weaken admission controls and deploy workloads that PSS policies would otherwise block. Important: The final …

EasyAdminBundle ships two public Twig components — <twig:ea:Flag countryCode="…"> and <twig:ea:Icon name="…"> — that load SVG files from disk using a path built directly from a public component property, and then render the resulting markup with the Twig |raw filter. When an application binds either of those properties to data that is influenced by an end user, the lack of validation on the property value leads to two distinct issues: …

CrateDB has two ways to access blob storage: SQL (SELECT … FROM blob.<table> and friends) and the blob HTTP API (GET|PUT|DELETE /_blobs/{table}/{digest}). The SQL path goes through AccessControl, which is what enforces privilege grants; that's why SELECT digest FROM blob.secret_blobs fails for a user who has no grants on the table. The HTTP path authenticates the request but never asks AccessControl whether the authenticated user is allowed to touch the …

The Cortex MCP server (neuro-cortex-memory) treats the CLAUDE_PROJECT_DIR environment variable — automatically set by Claude Code to the currently open project directory — as a trusted Cortex developer checkout. When the open_visualization tool is invoked, _find_dev_source() resolves the user's active project directory as a candidate Cortex source root. The only validation performed by _is_cortex_root() is a check for the presence of an mcp_server/ subdirectory and a ui/unified-viz.html file. An attacker …

Config.registryFor selected a per-registry credential / CA / mirror block by checking strings.HasSuffix(name, fqdn) after stripping a single trailing dot. The match has no boundary between the configured FQDN and any preceding characters in the request hostname. A registry configured as [registries."ghcr.io."] is therefore also applied to any image pulled from a host whose name happens to end in the literal byte sequence ghcr.io, including attacker-registered domains such as evilghcr.io. …

ciphertextContainer.UnmarshalJSON decodes the third :-separated component of a vault:vX:base64… ciphertext and then unconditionally takes a 12-byte prefix slice for the AES-GCM nonce: c.nonce = fullCiphertext[:aesGCMNonceSize]. If the decoded blob is shorter than 12 bytes, the slice expression panics. The panic happens before any cryptographic operation, while the JSON body of the request is still being parsed inside the request handler. Because the handler is invoked from net/http's standard handler goroutine, …

An attacker is able to craft and send a user a URL that will redirect the user from the Concourse web server to any other site. This could be used in a phishing attack to steal user's credentials.

Centrifugo's dynamic JWKS endpoint feature can verify a JWT for one allowed issuer using a public key cached from another allowed issuer. The JWKS cache and singleflight lookup are keyed only by the JWT header kid, not by the resolved JWKS endpoint, issuer, audience, or other trust-domain namespace. In a documented multi-issuer dynamic JWKS configuration, an attacker who can obtain or mint a valid token for issuer/tenant A can authenticate …

Centrifugo's dynamic JWKS endpoint feature can verify a JWT for one allowed issuer using a public key cached from another allowed issuer. The JWKS cache and singleflight lookup are keyed only by the JWT header kid, not by the resolved JWKS endpoint, issuer, audience, or other trust-domain namespace. In a documented multi-issuer dynamic JWKS configuration, an attacker who can obtain or mint a valid token for issuer/tenant A can authenticate …

Centrifugo's dynamic JWKS endpoint feature can verify a JWT for one allowed issuer using a public key cached from another allowed issuer. The JWKS cache and singleflight lookup are keyed only by the JWT header kid, not by the resolved JWKS endpoint, issuer, audience, or other trust-domain namespace. In a documented multi-issuer dynamic JWKS configuration, an attacker who can obtain or mint a valid token for issuer/tenant A can authenticate …

Centrifugo's dynamic JWKS endpoint feature can verify a JWT for one allowed issuer using a public key cached from another allowed issuer. The JWKS cache and singleflight lookup are keyed only by the JWT header kid, not by the resolved JWKS endpoint, issuer, audience, or other trust-domain namespace. In a documented multi-issuer dynamic JWKS configuration, an attacker who can obtain or mint a valid token for issuer/tenant A can authenticate …

Centrifugo's dynamic JWKS endpoint feature can verify a JWT for one allowed issuer using a public key cached from another allowed issuer. The JWKS cache and singleflight lookup are keyed only by the JWT header kid, not by the resolved JWKS endpoint, issuer, audience, or other trust-domain namespace. In a documented multi-issuer dynamic JWKS configuration, an attacker who can obtain or mint a valid token for issuer/tenant A can authenticate …

auth-fetch-mcp v3.0.1 implements SSRF protection in assertSafeUrl() (src/security.ts) to block requests to private and loopback addresses. However, the isPrivateV6() function fails to detect IPv4-mapped IPv6 loopback addresses in their hex-normalized form. When an attacker supplies a URL such as http://[::ffff:127.0.0.1]:PORT/, the Node.js WHATWG URL parser silently normalizes the host to [::ffff:7f00:1]. Because net.isIPv4('7f00:1') returns false, the private-IP check is bypassed and the URL is passed to the browser or HTTP …

@apify/actors-mcp-server version 0.10.7 builds Actor standby URLs by directly concatenating a trusted base URL with an attacker-controlled webServerMcpPath value taken from an Actor definition returned by the Apify API. An attacker who publishes a malicious Actor with a crafted webServerMcpPath (e.g., @attacker.example/mcp) can cause the MCP client to resolve the final URL to an entirely different host. Because the MCP client unconditionally attaches the victim's Authorization: Bearer <APIFY_TOKEN> header to …

Primary: arbitrary-CWD-file read primitive. An attacker-controlled OCI artifact, when pulled by a victim using the oras CLI or any Go program using oras-go/v2/content/file, can create a hardlink inside the victim's extract tree pointing to an arbitrary file in the victim's process CWD (that the invoker UID is permitted to read). Reading the extract-tree hardlink yields that file's contents verbatim. Secondary: inode-sharing tampering primitive. Any tool that later modifies the extract-tree …

The network domain has a central SSRF authorization policy that blocks private, loopback, link-local, and reserved targets unless an explicit authorization object allows private network access. The policy is enforced by raw HTTP/TCP/TLS RTT tools, but the ICMP probe and traceroute tools resolve the target and invoke the native ICMP/traceroute sink directly. An MCP client with access to an active network domain can therefore ask the jshookmcp server to probe …

dist/clients/core/params.ts in @hey-api/openapi-ts ships a runtime template that is copied verbatim into every generated SDK as params.gen.ts. When a caller passes an object argument containing an unknown key starting with a slot prefix ($body_, $headers_, $path_, $query_), the function strips the prefix and writes the remainder directly to that slot without validation. The key "$query___proto__" causes the returned params.query object to have its prototype chain substituted with attacker-controlled data. The …

The 3.26.0 source-policy hardening changed the signature of CoreExtension::checkArrow() to take a boolean $isSandboxed instead of an Environment, and added the same $isSandboxed argument to CoreExtension::arraySome() and CoreExtension::arrayEvery(). Compiled templates were updated to pass the per-source sandbox state computed at the call site. The deprecated internal wrappers exposed in src/Resources/core.php for legacy third-party code (twig_check_arrow_in_sandbox(), twig_array_some(), twig_array_every()) were not updated: twig_array_some() and twig_array_every() call CoreExtension::arraySome() / arrayEvery() without forwarding the …

This is a residual bypass of CVE-2026-46635 / GHSA-vcc8-phrv-43wj that only affects sandboxing enabled through SourcePolicyInterface (and not the regular global sandbox mode). CoreExtension::column() receives the active sandbox state via the needs_is_sandboxed channel as a boolean $isSandboxed, but then routes the per-element property reads through SandboxExtension::checkPropertyAllowed() without forwarding the current Source. SandboxExtension::checkPropertyAllowed() re-evaluates isSandboxed($source) internally; with $source = null the SourcePolicyInterface-driven decision is lost, the method short-circuits to "not sandboxed", …

This is a residual bypass of CVE-2026-47732 / GHSA-pr2w-4gpj-cpq4 left after the initial fix for unguarded __toString() calls. In 3.26.0 the sandbox visitor was extended to wrap every child node that its parent will string-coerce at runtime with CheckToStringNode, gated by the new CoercesChildrenToStringInterface. ArrayExpression did not implement the interface for its mapping keys: when a dynamic key expression resolves to a Stringable object, ArrayExpression::compile() emits a raw (string) cast …

This is a residual bypass of CVE-2026-47732 / GHSA-pr2w-4gpj-cpq4 left after the initial fix for unguarded __toString() calls. It covers two related coercion points that were not caught by the original patch. Traversable in join and replace filters. SandboxExtension::ensureToStringAllowed() recurses into PHP arrays so that a Stringable object hidden inside an array argument cannot be string-coerced without consulting the security policy. The recursion stops at PHP arrays: a Traversable value …

An unauthenticated remote attacker can trigger unbounded memory growth on the timestamp authority server. This vulnerability exists because the global wrapMetrics middleware records the raw HTTP request path (r.URL.Path) and raw HTTP request method (r.Method) as Prometheus labels for latency and request count metric vectors. Since this middleware runs before standard routing occurs, it executes for all incoming requests, including those for unmatched paths (yielding 404 responses) or arbitrary request …

An unauthenticated remote attacker can trigger unbounded memory growth on the timestamp authority server. This vulnerability exists because the global wrapMetrics middleware records the raw HTTP request path (r.URL.Path) and raw HTTP request method (r.Method) as Prometheus labels for latency and request count metric vectors. Since this middleware runs before standard routing occurs, it executes for all incoming requests, including those for unmatched paths (yielding 404 responses) or arbitrary request …

Regression: Verification of integratedTime from Rekor V1 Log Entry against Fuclio Certificate validity was missing

Probo's saferedirect package validates redirect URLs used across authentication flows (OIDC, SAML, session transfer, OAuth connectors, and trust-center magic links). The validator only inspected the second character of relative paths, so a URL like /../\evil.com passed validation because the second character is .. Go's http.Redirect normalizes this path to /\evil.com before setting the Location header. Browsers can interpret the backslash as a host separator and redirect the user to an …

The checkout component improperly filters URL-writable properties, allowing authenticated users to inject arbitrary key-value pairs into server provisioning parameters. Because bundled server extensions prioritize these user-supplied properties over administrator-defined configurations, a regular user can override hosting plans and resource limits at checkout without special privileges.

The credit payment implementation in app/Livewire/Invoices/Show.php executes a pessimistic row lock (lockForUpdate()) outside of an active database transaction. Because MySQL/MariaDB requires an enclosing transaction to enforce row-level locks, the guard is ineffective. Concurrent payment requests can exploit this race condition to read the same credit balance simultaneously, allowing users to pay multiple invoices using the same credit balance.

A memory-safety vulnerability in Open Babel's GAMESS output parser caused a use-after-free when reading a crafted input file.

A memory-safety vulnerability in Open Babel's bundled zipstream decompression code caused an out-of-bounds write via overlapping memcpy when reading a crafted gzip-compressed chemistry file.

A memory-safety vulnerability in Open Babel's MOL2 file format parser caused a NULL pointer dereference when reading a crafted input file.

A memory-safety vulnerability in Open Babel's SMILES parser caused a heap buffer overflow when reading a crafted input string.

Summary A memory-safety vulnerability in Open Babel's CIF file format parser allowed an out-of-bounds read when reading a crafted input file. Details The flaw was in OpenBabel::transform3d::DescribeAsString. A malformed symmetry-operation string caused the parser to read past the end of its internal buffer while formatting the description. Impact Open Babel is a C++ library and CLI used to read and write chemistry file formats; it is shipped by Linux distributions …

A memory-safety vulnerability in Open Babel's CDXML file format parser caused a NULL pointer dereference when reading a crafted input file.

A small OpenAPI document containing a circular schema reference can cause process termination through stack overflow in Microsoft.OpenApi. The issue affects OpenAPI document parsing through public OpenAPI.NET reader APIs and has been confirmed across both JSON and YAML reader paths. Applications, CLIs, developer tools, or services that parse untrusted OpenAPI documents in-process may be terminated by a crafted OpenAPI document containing circular schema references. The impact is availability/process termination only. …

Kahi releases up to and including v0.1.0-alpha.8 contain three privilege/permission issues, all fixed in v0.1.0-alpha.9. They were identified in a full-codebase security review on 2026-05-26.

Three security vulnerabilities were identified in the OIDC Discovery client: Blind Server-Side Request Forgery (SSRF) via Cross-Host Redirects: Fulcio uses an HTTP client to fetch OIDC discovery metadata (/.well-known/openid-configuration). Prior to this fix, if a configured issuer returned an HTTP redirect to a different host, the client followed it by default. This allowed a compromised or malicious issuer to redirect Fulcio's discovery requests to internal-only systems, resulting in blind SSRF. …

The Fission MessageQueueTrigger (MQT) scaler controller exposed two privilege-escalation primitives to any subject able to create MQTs in a namespace.

A follow-up bypass of the round-4 PodSpec hardening (GHSA-gx55-f84r-v3r7, GHSA-wmgg-3p4h-48x7, GHSA-v455-mv2v-5g92). Those advisories validate and sanitize the PodSpec (spec.runtime.podSpec / spec.builder.podSpec / function.spec.podSpec), but the Environment CRD also exposes spec.runtime.container and spec.builder.container — a standalone Container merged into the runtime/builder pod whose SecurityContext bypassed both layers.

A Fission Function spec carries three reference types — Secret, ConfigMap, and Package. The first two were namespace-validated by the admission webhook; PackageRef.Namespace was not.

A low-privilege developer who could create a KubernetesWatchTrigger (KWT) in their own namespace was able to establish a persistent surveillance channel over any other namespace.

The Fission Function admission webhook (pkg/webhook/function.go) validated that spec.secrets[].namespace and spec.configmaps[].namespace equalled the function's own namespace but performed no equivalent check on spec.environment.namespace.

Fission's buildermgr controller processed Package CRDs without verifying that Package.spec.environment.namespace matched Package.metadata.namespace.

Fission's Environment CRD exposes spec.runtime.podSpec and spec.builder.podSpec, which are merged into the Kubernetes pod specs for runtime and builder pods. The merge logic propagated hostNetwork, hostPID, hostIPC, container privileged, and serviceAccountName from the user-supplied podspec with no filtering, and Environment.Validate performed no security-relevant checks on these fields.

A stronger framing of the same root cause as GHSA-gx55-f84r-v3r7: the Environment.spec.runtime.podSpec / spec.builder.podSpec passthrough lacked validation, and MergePodSpec propagated dangerous fields into the generated pods.

Fission's Container Executor path lets a tenant supply Function.spec.podspec directly; the executor merges it into the executor-built podspec and creates a Deployment whose pods run the user's container image.

Fission builder pods were created with ServiceAccountName: fission-builder and no AutomountServiceAccountToken: false, so the kubelet auto-mounted the service-account token into every container in the pod — including the user-supplied builder image.

FolderSchemeHandlerFactory was intended to restrict served files to a configured rootFolder, but its path validation used a raw string prefix check. A request could escape to a sibling directory whose full path starts with the root folder path, allowing files outside the configured root to be served.

@cedar-policy/authorization-for-expressjs is an open-source Express.js middleware that integrates Cedar authorization into Express applications by mapping HTTP requests to Cedar actions and evaluating authorization policies before allowing requests to proceed. An issue exists where, under certain circumstances, the middleware matches incoming requests against Cedar action mappings using req.originalUrl, which includes the query string, while Express routes requests using only the path component.

The fix for GHSA-f5x2-vj4h-vg4c / CVE-2026-25754 introduced in commit 40e1c71 is incomplete and can be bypassed through nested prototype pollution payloads. The original patch replaced the internal FormFields storage object with Object.create(null), preventing direct payloads such as proto.polluted. However, payloads containing a non-dangerous segment before proto or constructor.prototype, such as user.proto.polluted, still lead to Object.prototype pollution. This issue is exploitable remotely through a single unauthenticated multipart/form-data request using the default …

Description An Improper Authentication (CWE-287) issue in OpenAM's OAuth2 private_key_jwt client authentication path allows any registered OAuth2 client to mint tokens in the name of any other client whose key is published via a jwks_uri, without knowing the victim's signing key. This affects OpenAM Community Edition through version 16.0.6 and was patched in version 16.1.1.

Description An Improper Authorization (CWE-285) issue in OpenAM's OAuth2 authorization-code grant allows a PKCE-protected authorization code to be redeemed without the required code_verifier. This affects OpenAM Community Edition through version 16.0.6 and was patched in version 16.1.1. The authorize endpoint stores a code_challenge on the issued code, but the token endpoint only requires a code_verifier when the realm-wide codeVerifierEnforced setting is enabled, which ships disabled by default. With that setting …

Description A Protection Mechanism Failure (CWE-693) in OpenAM's server-side scripting sandbox allows an authenticated script author execute operating-system commands from the OpenAM JVM with the default class allow and deny lists. This impacts OpenAM Community Edition through version 16.0.6. This issue was patched in version 16.1.1.

The checkUserPassword GraphQL query in Dgraph is vulnerable to DQL (Dgraph Query Language) injection. User-supplied password values are interpolated directly into a DQL checkpwd() query via fmt.Sprintf without any escaping or parameterization. An attacker can inject a password containing a double-quote character to break out of the DQL string literal and append arbitrary DQL query blocks.

pnpm accepts package names from the env lockfile configDependencies section and uses those names directly when creating config dependency symlinks under node_modules/.pnpm-config. A malicious repository can commit a crafted pnpm-lock.yaml whose env-lockfile document contains a traversal-shaped config dependency name such as ../../PWNED_CFGDEP. During pnpm install, pnpm installs the config dependency and creates a symlink at a path derived from that name. In local testing against pnpm v11.5.1, this caused pnpm …

The hoisted dependency alias issue tracked as GHSA-fr4h-3cph-29xv / CAND-PNPM-059 has been addressed in both pnpm and pacquet. A crafted lockfile alias could be joined directly under a hoisted node_modules directory. Traversal aliases could escape that directory, while reserved aliases such as .bin or .pnpm could overwrite pnpm-owned layout. This patch validates package-name semantics and path containment before graph insertion or filesystem work.

The patch-remove deletion-scope issue tracked as GHSA-72r4-9c5j-mj57 / CAND-PNPM-030 has been addressed in pnpm. A crafted patch entry could resolve outside the configured patches directory and cause pnpm patch-remove to delete an arbitrary reachable file. This patch validates the configured directory and every resolved target before unlinking anything, then deletes the final directory entry without following it.

YARD's static cache lookup reads a request path before the router's path cleanup runs. When a server is configured with a document root, a traversal path such as /../yard-cache-secret.html is joined against that root and can return a readable sibling .html file outside the intended static tree. The potential security risk seems low, as only html-ending files can be read, but still the risk of reading arbitrary html files is …

Webauthn\Bundle\Security\Http\Authenticator\WebauthnAuthenticator logs the full Symfony\Component\HttpFoundation\Request object inside the log context of both onAuthenticationSuccess() and onAuthenticationFailure() at INFO level: $this->logger->info('User has been authenticated successfully with Webauthn.', [ 'request' => $request, 'firewallName' => $firewallName, 'identifier' => $token->getUserIdentifier(), ]); $this->logger->info('Webauthn authentication request failed.', [ 'request' => $request, 'exception' => $exception, ]); Request::__toString() returns the raw HTTP message, including every request header. As soon as the configured logger normalises or stringifies the context (default …

turso-cli persists the user's Turso platform JWT to settings.json using Viper's default configPermissions of 0o644, leaving the credential file world-readable on standard Linux and macOS systems. Any other local UID on the host can read the file and recover the platform JWT, which grants full Turso platform access scoped to the user's organizations.

In gonic, the Subsonic API endpoints /rest/deletePlaylist.view and /rest/getPlaylist.view perform no per-resource authorization. Once authenticated as any user (admin or not), an attacker can: Delete any playlist owned by any other user (including admin) by passing its id. Read the full contents (name, comment, song list) of any other user's private (non-public) playlist by passing its id. The Subsonic playlist id is base64url("<userID>/<filename>.m3u"). Because filenames are user-supplied or time-derived and …

line-desktop-mcp supports a –http-mode Streamable HTTP transport for use with clients such as n8n. In this mode the server binds to 0.0.0.0 and exposes the MCP /mcp endpoint without an MCP-layer authentication check. Any network client that can reach the port can initialize a session, list tools, and call tools that read LINE Desktop chat history or send LINE messages through the already logged-in desktop application. This is High for …

The Glide image proxy's URL validation could be bypassed using DNS rebinding. The remote hostname was validated as publicly routable, but resolved again when the image was actually fetched, so an attacker controlling the hostname's DNS could rebind it to an internal address after validation. This could cause the server to make HTTP requests to internal addresses — including loopback, private network, and cloud metadata endpoints. This affects sites that …

Form submission values were not neutralized for spreadsheet formula characters when exported to CSV. A submission containing a value beginning with a formula trigger character (e.g. = , + , - , @ ) could be interpreted as a live formula when a Control Panel user opens the export in a spreadsheet application. Form submissions can come from unauthenticated front-end visitors, so the malicious value can be supplied by an …

The fix for GHSA-4jjr-vmv7-wh4w was incomplete. It addressed the issue in the query builder, but the same protection was not applied to in-memory collection sorting. Manipulating sort parameters could result in the loss of content and assets. This requires a front-end template that passes request input into a tag's sort parameter. It is not exploitable by default — a template would need to be explicitly set up to sort by …

The Live Preview endpoint for existing entries and terms only checked view authorization, but it accepts and renders caller-supplied field values. A Control Panel user with view but not edit permission could therefore submit content they were not authorized to author and generate a shareable Live Preview URL rendering it.

An authenticated Control Panel user could view metadata and content for resources they don't have permission to view, including entries, assets, users, roles, groups, and other configured resources. Depending on the resource, this could expose titles, custom field values, entry content, asset metadata, and the existence of users, roles, and groups. No data could be modified.

Four authorization bypass vulnerabilities in Symfony LiveComponent actions allow any authenticated user within a company to access, modify, or delete other users' API tokens and notification transport settings. The root cause is that LiveComponent actions accept entity IDs without verifying ownership, while the listing methods correctly filter by user.

semantic-router versions 0.1.8 through 0.1.14 declare litellm>=1.61.3 with no upper bound. During the window in which litellm==1.82.8 was the latest release on PyPI, a fresh install of any affected semantic-router version could resolve to that compromised wheel. The malicious litellm==1.82.8 wheel ships a litellm_init.pth file that executes on Python interpreter startup — no import required. It collects and exfiltrates: Process environment variables AWS / GCP / Azure credentials SSH keys, …

The ExpressionDepthLimit parser guard in Scriban does not actually stop parsing — it only logs a non-fatal error and lets recursive descent continue. As a result, a template containing a deeply nested expression (parentheses, array initializers, object initializers, or unary operators) drives the recursive-descent parser into a native stack overflow. The resulting StackOverflowException is uncatchable in .NET and immediately terminates the host process. Any application that parses an attacker-influenced template …

The array multiplication operator (array * integer) in Scriban allocates a result whose size is the product of the attacker-controlled integer and the array length, with no LoopLimit / LimitToString check and no overflow-safe arithmetic. A ~40-byte template forces a multi-gigabyte allocation, producing a denial-of-service. This is the unguarded sibling of operations that were hardened against the same class of abuse: string * integer (gated by a LimitToString pre-check), array.insert_at …

The remark42 image proxy fetches an arbitrary remote URL and re-serves the response from remark42's own origin. The download path decides whether the fetched resource is an image by looking only at the Content-Type header the remote server claims — it never inspects the actual bytes. The serving path then derives the response Content-Type by sniffing those bytes with http.DetectContentType. An attacker hosts a URL that sets Content-Type to image/png …

Credentials for a registry may be inadvertently leaked to external servers. A prerequisite for this attack is a malicious registry server, a malicious blob store, or a registry that does not restrict the external URLs for foreign blobs.

The python-socketio server stores binary EVENT and ACK messages in memory while it waits to receive their binary attachments. Once all the attachments are received, these messages are then processed. An attacker can submit a binary message and intentionally omit sending one or more of its attachments to cause the message along with the partial list of received attachments to stay in memory for a long time.

An attacker can cause the creation of unnecessary background threads in the python-engineio server by exploiting the heartbeat mechanism, which launches a thread when a new connection is received, and when the client sends a PONG packet. Note: this issue primarily affects synchronous servers. Asynchronous servers allocate background tasks instead of physical threads, which are lightweight and less likely to cause denial of service. However, the fix that was implemented …

There are two specific configurations of the python-engineio server in which the size of incoming messages is not checked before the messages are loaded into memory. An attacker can take advantage of these to cause unnecessary memory allocations in the python-engineio server. The two cases are: POST requests, when using ASGI with the long polling transport WebSocket messages, when using Aiohttp with the WebSocket transport

When an application using Pydantic AI opts a URL into force_download='allow-local' (which disables the default block on private/internal IPs) and runs on a network that routes the affected IPv6 transition forms (NAT64- or ISATAP-configured networks), the cloud-metadata blocklist could be bypassed by encoding the metadata IP in an IPv6 transition form that the previous fix did not decode — IPv4-compatible IPv6 (::a.b.c.d), the NAT64 RFC 8215 local-use prefix (64:ff9b:1::/48), operator-chosen …

When an application using Pydantic AI opts a URL into force_download='allow-local' (which disables the default block on private/internal IPs) and runs on a network that routes the affected IPv6 transition forms (NAT64- or ISATAP-configured networks), the cloud-metadata blocklist could be bypassed by encoding the metadata IP in an IPv6 transition form that the previous fix did not decode — IPv4-compatible IPv6 (::a.b.c.d), the NAT64 RFC 8215 local-use prefix (64:ff9b:1::/48), operator-chosen …

In wings/internal/ufs/fs_unix.go (line 92-94), this function is defined and is used to change permissions of files in the server: func (fs UnixFS) fchmodat(op string, dirfd int, name string, mode FileMode) error { return ensurePathError(unix.Fchmodat(dirfd, name, uint32(mode), 0), op, name) } This call to the unix function fchmodat(int fd, char name, mode_t mode, int flags) does not have the flag AT_SYMLINK_NOFOLLOW set, and Wings neither checks or validate if the target …

An unprotected user enumeration vulnerability exists in the account email update endpoint, allowing authenticated users to verify whether email addresses are registered on the panel through automated requests without rate limiting or CAPTCHA protection.

pnpm install in non-frozen mode can accept new remote package content after detecting that the downloaded tarball does not match the integrity recorded in pnpm-lock.yaml. When a package is already locked with an integrity value, and the registry later serves different metadata and tarball content for the same package name and version, pnpm initially reports an integrity mismatch. However, plain pnpm install then performs a resolution repair, accepts the registry's …

pnpm allows a transitive dependency alias from registry package metadata to contain path traversal segments. During install, pnpm later uses that alias as a filesystem path when linking dependency nodes. As a result, a registry package can cause pnpm install - ignore-scripts to replace paths in the current project with symlinks to attacker-controlled dependency package directories. .git/hooks is only one useful target. The same primitive can replace other project-local paths …

A malicious codeload.github.com server can serve whatever tarball it wants and pnpm will install it regardless of the lockfile.

Manifest bin object keys such as "", ".", and ".." passed pnpm's bin-name guard. When a malicious package was installed globally, later global remove, update, or add-replacement flows could re-derive those names from the installed manifest and pass path.join(globalBinDir, binName) to removeBin. For "." this targets the global bin directory; for ".." this targets its parent.

pnpm can install configDependencies declared in pnpm-workspace.yaml before command dispatch. Before the patch, a repository could declare pacquet or @pnpm/pacquet as a config dependency and pnpm treated that repository-controlled dependency as an install-engine opt-in. During install, pnpm resolved a platform-specific @pacquet/<platform>-<arch>/pacquet binary from node_modules/.pnpm-config/<packageName> and spawned it as the developer or CI user.

pnpm and pacquet expanded ${ENV_VAR} placeholders from repository-controlled .npmrc and pnpm-workspace.yaml into registry request destinations and registry credentials. A malicious repository could cause dependency resolution to send victim environment secrets to an attacker-selected registry before lifecycle scripts run.

pnpm can persist package-manager bootstrap metadata in the first YAML document of pnpm-lock.yaml. Before the patch, direct pnpm execution trusted an already resolved packageManagerDependencies entry when the committed env lockfile contained matching pnpm and @pnpm/exe versions. A malicious repository could therefore commit package-manager lockfile package records and snapshots that bypassed fresh package-manager resolution, then cause pnpm to install and execute bytes selected by that committed lockfile state during automatic version …

Keep build approval for opaque dependency sources byte-exact for GHSA-5wx6-mg75-v57r / CAND-PNPM-123. Merged upstream commit bf1b731ee6 fixed the original name-only approval bypass by making build policy consume the resolved dependency identity. One collision remained: the generic peer-suffix normalizer also stripped parenthesized text from git, URL, tarball, file, and other opaque locators. Approval for one source string could therefore authorize a different attacker-controlled source whose locator normalized to the same value.

pnpm passes the lockfile-controlled git resolution.commit value to git fetch without a – separator or commit-format validation. For git dependencies fetched through the shallow-fetch path, a malicious lockfile can replace the expected 40-character commit hash with a Git option such as –upload-pack=<command>. For SSH and local transports, –upload-pack can execute the supplied command. HTTPS transports ignore –upload-pack, so the practical attack surface is primarily SSH or local git dependencies.

The staged-tarball filename traversal reported as GHSA-v23m-ccfg-pq9h / CAND-PNPM-038 is fixed on main by pnpm/pnpm#12303, merged as 65443f4bdf1f0db9c8c7dc58fee25252607e9234. Before the fix, pnpm stage download derived a local filename from registry-controlled package name and version fields. A crafted manifest could escape the selected download directory and overwrite another reachable file. The merged fix validates both fields, derives one safe filename, and verifies the final destination before writing.

pnpm's patch application pipeline (@pnpm/patch-package) performs no path validation on file paths extracted from .patch files. An attacker who contributes a malicious patch file via a pull request can write attacker-controlled content to or delete arbitrary files on the filesystem during pnpm install, as the user running the install. The diff –git header paths containing ../../ sequences traverse out of the package directory, and the traversal is difficult to catch …

pnpm's tarball extraction worker skips integrity verification when the integrity field is absent from the lockfile resolution. If an attacker can both modify pnpm-lock.yaml to remove the integrity: field and cause the referenced registry URL to serve altered package content, pnpm install –frozen-lockfile can install the altered package without an integrity error. npm's npm ci enforces integrity by default; pnpm's behavior of silently skipping verification is a pnpm-specific fail-open gap.

pnpm can send user-level unscoped npm authentication credentials to a registry chosen by a repository-local .npmrc file. In the reproduced case, the user's npm config contains a default registry and an unscoped _authToken. The repository does not provide a token-bearing auth line. It only sets registry= to a different registry URL. During normal pnpm metadata/install workflows, pnpm binds the user-origin unscoped credential to the repository-selected registry and sends it as …

pontedilana/php-weasyprint fetches the content of option values server-side via file_get_contents() when the value looks like a URL, without restricting the URL scheme. The attachment option of Pdf is the reachable sink: any value that passes isOptionUrl() (filter_var(…, FILTER_VALIDATE_URL)) is downloaded by the PHP process and embedded into the generated PDF. Because FILTER_VALIDATE_URL accepts http, https, ftp, file and PHP stream wrappers such as php://, an attacker who can influence the …

pontedilana/php-weasyprint guarded the output filename against the phar:// stream wrapper with a case-sensitive blacklist: if (0 === \strpos($filename, 'phar://')) { throw new \InvalidArgumentException('The output file cannot be a phar archive.'); } PHP stream wrappers are case-insensitive, so PHAR://, Phar://, etc. bypass the check and reach fileExists() (file_exists()) in prepareOutput(). On PHP 7 (which the library still supports — PHP 7.4+), this triggers deserialization of a crafted PHAR archive's metadata, leading …

AbstractGenerator::$temporaryFiles is a public array, and removeTemporaryFiles() — invoked from __destruct() and from a registered shutdown function — calls unlink() on every entry without verifying that the path is contained within the temporary folder. Any code holding a reference to a generator instance can push an arbitrary path into the array and have it deleted on script shutdown. This mirrors the KnpLabs/snappy issue GHSA-87qc-37cw-84h4, patched in snappy 1.7.2.

phpMyFAQ 4.1.3 fixed GHSA-xvp4-phqj-cjr3 ("IDOR Account Takeover") by adding actor-authorization guards to UserController::overwritePassword(). The patch establishes a new invariant, stated in its own code comments: "Only SuperAdmins may change other users' [attributes]. Self-service is always allowed." and "a non-SuperAdmin must never be able to alter a SuperAdmin or protected account." That invariant is not enforced on two sibling endpoints in the same file, which the 4.1.3 fix left unchanged, and …

phpMyFAQ 4.1.3 fixed GHSA-xvp4-phqj-cjr3 ("IDOR Account Takeover") by adding actor-authorization guards to UserController::overwritePassword(). The patch establishes a new invariant, stated in its own code comments: "Only SuperAdmins may change other users' [attributes]. Self-service is always allowed." and "a non-SuperAdmin must never be able to alter a SuperAdmin or protected account." That invariant is not enforced on two sibling endpoints in the same file, which the 4.1.3 fix left unchanged, and …

pontedilana/php-weasyprint builds the shell command for WeasyPrint by passing the binary path through escapeshellarg() first and then checking the quoted result with is_executable(). On POSIX escapeshellarg('/usr/local/bin/weasyprint') returns '/usr/local/bin/weasyprint' with the single-quote characters as part of the string, so is_executable() looks for a file whose actual name includes those quotes. That file never exists, the "safe" branch is dead code, and the raw $binary string (set via the constructor or setBinary()) …

Psl\H2\ServerConnection does not validate that the total bytes received in DATA frames match the content-length header declared in the HEADERS frame, in violation of RFC 9113 §8.1.1. A malicious client can: Send more DATA bytes than declared, smuggling additional content past application-level size limits. Send fewer DATA bytes than declared and close the stream early, causing applications that trust the declared length to behave incorrectly. The vulnerability is only reachable …

Psl\H2\ServerConnection does not validate that the total bytes received in DATA frames match the content-length header declared in the HEADERS frame, in violation of RFC 9113 §8.1.1. A malicious client can: Send more DATA bytes than declared, smuggling additional content past application-level size limits. Send fewer DATA bytes than declared and close the stream early, causing applications that trust the declared length to behave incorrectly. The vulnerability is only reachable …

Description An LDAP Injection (CWE-90) vulnerability in the MSISDN authentication module allows an unauthenticated, remote attacker to obtain an arbitrary OpenAM session without a password in the default trusted gateway configuration. This impacts OpenAM Community Edition through version 16.0.6. This issue was patched in version 16.1.1.

Description An Unverified Password Change (CWE-620) and Use of Weak Credentials (CWE-1391) issue in OpenAM's OAuth2 authentication module silently rewrites a local user's password to the literal string of their username on OAuth2 re-login of an existing account. The default ldapService chain then accepts the username as the password for that user, allowing an unauthenticated attacker to obtain a session via the standard authenticate endpoint with both username and password …

nono-py policy handling could fail open in two ways. First, resolving a policy-derived ProxyConfig did not automatically enforce CapabilitySet.proxy_only, allowing sandboxed children to bypass a resolved domain allowlist by using direct network access. Second, policy JSON accepted unknown security-sensitive fields, so misspelled or unsupported restrictions could be silently ignored.

The python API made a restrictive-looking configuration unsafe by default. A caller could configure only reverse- proxy credential routes, put the child in CapabilitySet.proxy_only, and reasonably expect network access to be limited to those routes. Instead, because empty allowed_hosts meant allow-all inside nono-proxy, the child could use the local proxy as a transparent CONNECT tunnel to non-route nominated hosts (not including metadata endpoints). That is an authorization bypass / policy …

On Linux kernels that do not support Landlock network rules, nono_py.sandboxed_exec() could run CapabilitySet.proxy_only(proxy) without supervising the seccomp-notify proxy-only fallback returned by the Rust core. In that configuration, a sandboxed child process could remove HTTP_PROXY / HTTPS_PROXY environment variables or use raw sockets and then open direct TCP connections that should have been denied by proxy-only policy. The issue affects proxy-only enforcement. It does not mean that all nono-py network …

In nezha v1.14.13–v1.14.14 and v2.0.0–v2.0.9, the WebSocket endpoints GET /ws/terminal/:id and GET /ws/file/:id authenticate the caller only by the presence of a valid stream UUID, with no ownership check tying that UUID to the user who created the stream. Any authenticated dashboard user (including a RoleMember) who learns a live stream UUID can attach to the session and gain interactive shell access or full file-manager control on the target server …

1. Description The Nezha dashboard exposes two endpoints that create long-lived WebSocket streams to monitored agents: POST /api/v1/terminal → createTerminal() (terminal.go:27-67) POST /api/v1/file → createFM() (fm.go:28-67) Both call rpc.NezhaHandlerSingleton.CreateStream(streamId, …) which inserts a new ioStreamContext into an unbounded map[string]*ioStreamContext (s.ioStreams in io_stream.go:59-67). There is no per-user rate limit, no global semaphore, and no per-server connection cap. Each stream allocates: A ioStreamContext struct with several channels and sync primitives Two goroutines …

PATCH /server/{id} accepts and persists nonexistent ddns_profiles IDs for a member-owned server. If another user later creates a DDNS profile with one of those IDs, the DDNS worker resolves the stored ID and dispatches an update using the other user's DDNS profile configuration in the context of the attacker's server. This is a second-order authorization bypass: direct binding to an existing foreign DDNS profile is correctly denied, but an unresolved …

fallbackToFrontend in the dashboard's NoRoute handler treats any URL whose raw string starts with /dashboard as an admin-frontend asset request. The check uses strings.HasPrefix, not a path-segment match, so the input /dashboard../data/config.yaml is accepted; strings.TrimPrefix leaves ../data/config.yaml; and path.Join("admin-dist", "../data/config.yaml") normalizes to data/config.yaml — which os.Stat finds and http.ServeFile returns. No authentication required. In default deployments (the values shipped in model/config.go and the layout shipped in the project Dockerfile) data/config.yaml …

1. Description The getRedirectURL function in oauth2.go:22-29 constructs the OAuth2 callback URL by concatenating the request's Host header with a fixed path, with zero validation of the Host header: func getRedirectURL(c *gin.Context) string { scheme := "http://" referer := c.Request.Referer() if forwardedProto := c.Request.Header.Get("X-Forwarded-Proto"); forwardedProto == "https" || strings.HasPrefix(referer, "https://") { scheme = "https://" } return scheme + c.Request.Host + "/api/v1/oauth2/callback" } File: cmd/dashboard/controller/oauth2.go:22-29 This function is called from oauth2redirect() …

An authenticated non-admin user who owns any server can create or update a NAT profile whose domain is equal to the dashboard's own HTTP Host (for example, dashboard.example:8008). The dashboard's top-level HTTP/gRPC multiplexer checks NATShared.GetNATConfigByDomain(r.Host) before dispatching requests to the dashboard API, frontend, or gRPC handler, so a member-controlled NAT profile for the dashboard Host takes precedence over the real dashboard. A disabled claimed NAT profile blocks matching dashboard requests …

The GET /api/v1/ddns and GET /api/v1/notification endpoints return full resource objects including plaintext third-party API credentials — Cloudflare API tokens, TencentCloud SecretKeys, Slack/Discord/Telegram webhook URLs with embedded bot tokens, and Authorization header values — without any field-level redaction. Any authenticated admin who calls these endpoints receives every stored credential in the system in a single API response. A compromised admin session or leaked PAT with nezha:ddns:read or nezha:notification:read scope exposes …

internal/api/pop/nonce.go:25,40,86 + internal/api/server.go:38 — the signed-poll nonce cache is an in-process LRU sized at 65,536 entries. internal/api/updates.go:31 sets pollClockSkew = 5 * time.Minute as the replay window.

The web UI (/ui/*) does not apply the per-operator CA scoping the JSON API received for GHSA-598g-h2vc-h5vg. Any authenticated non-admin operator (for example, one created via self-registration or OIDC) can access resources belonging to other operators.

A NULL pointer dereference vulnerability exists in PDFParser::CreateFilterForStream() when processing a PDF stream with /Filter /LZWDecode and a /DecodeParms dictionary that does not contain the EarlyChange key. This causes an access violation (0xC0000005) and crashes the process.

mcp-pinot v3.0.1 (and earlier) defaults to running an HTTP MCP server bound to 0.0.0.0:8080 with no authentication enabled. All MCP tools, including SQL query execution, schema creation, and table-config mutation, are reachable by any network-adjacent caller. The server proxies these calls using server-side Pinot credentials, producing a confused-deputy condition that yields full read/write access to the configured Pinot cluster.

The HTTP MCP JSON-RPC endpoint at /mcp requires only OAuth read scope for all requests, then dispatches tools/call directly to handlers that include mutating tools. A read-only OAuth client can call store_memory and delete_memory through MCP even though the corresponding REST endpoints require write scope.

LinkifyIt.prototype.match — the package's primary public API — has O(N²) algorithmic complexity for inputs containing many fuzzy links or emails. This is not a regex backtrack bug; it's a structural issue in the JS-level scan loop that re-slices the input and re-runs unanchored regex searches on progressively shorter tails, N times. 64 KB of "a@b.com\n" repeated burns ~2.5 s of single-threaded CPU; 128 KB takes ~10 s. Doubling the input …

js-toml versions up to and including 1.1.0 parse hexadecimal / octal / binary integer literals via a hand-written parseBigInt loop that multiplies a BigInt accumulator by the radix once per input digit. Each iteration performs a BigInt * BigInt operation on an accumulator that grows linearly with the number of digits already consumed, so the whole loop is O(n²) in the literal length. The lexer regex places no upper bound …

js-toml's interpreter checks whether a key already exists in a parser-built container with if (object[key]) instead of if (key in object). When the prior value is a falsy primitive — false, 0, 0n, 0.0, -0, or "" — the duplicate-key branch is skipped and the value is silently overwritten by a later sub-table, dotted-key sub-table, or array-of-tables sharing the same name. Per the TOML 1.0.0 spec ("Defining a key multiple …

Testing revealed that joserfc accepts oversized RFC7797 b64=false JWS payloads without applying JWSRegistry.max_payload_length. The normal JWS compact and flattened JSON paths reject payloads above the configured payload-size limit with ExceededSizeError. The RFC7797 unencoded payload paths do not make the same check. A valid b64=false compact or flattened JSON JWS can therefore deserialize successfully with a payload larger than JWSRegistry.max_payload_length. This creates a moderate availability/resource-exhaustion risk for applications that accept lower-trust …

(backend).createDependentVolumesFromBackup in internal/server/storage/backend.go contains a cluster of unguarded pointer derefs on every dependent-volume entry's VolumeSnapshots[i], Volume, and Pool sub-fields. An authenticated user with can_create_instances permission on any project can crash the incusd daemon by uploading an instance backup tarball whose dependent_volumes[] block contains a nil snapshot pointer (or omits volume: / pool:). This is a sibling-field variant of the 2026-05-04 batch fix d768f81c0a1d985f35ae56219519822b080bf5e3 ("Properly check dependent volumes on import"). That …

(*backend).CreateCustomVolumeFromBackup in internal/server/storage/backend.go contains an unguarded time.Time dereference on the ExpiresAt field of every volume-snapshot entry in an imported custom-volume backup. An authenticated user with can_create_storage_volumes permission on any project can crash the incusd daemon by uploading a backup tarball whose volume_snapshots[].expires_at field is absent. This is a sibling-field variant of GHSA-r7w7-mmxr-47r9 (CVE-2026-40197). Commit 985a1dedf9f3e7ba729c93b654905ed510de25c2 added if s == nil at the top of the loop body, but did not …

A specially crafted image or instance backup can be used to read or create/write arbitrary files on the host; possibly leading to arbitrary command execution.

Improper validation of user-provided backup compression algorithm leads to argument injection in the constructed command line. This leads to an arbitrary file write on the host, possibly leading to arbitrary command execution.

The S3 protocol upload endpoint is vulnerable to path traversal and allows creation of arbitrary files on the host. This behavior could lead to arbitrary command execution.

An arbitrary file write exists in the Incus client when a malicious image server returns a crafted Incus-Image-Hash header. This can lead to arbitrary command execution as root on the server.

The record-output parameter of the /instances/$name/exec endpoint stores the output of the command in the exec-output directory of the instance. If exec-output is a symlink, file named exec_UUID.stdout and exec_UUID.stderr can be written to an arbitrary location where the .stdout file will contain arbitrary content. This behavior can be abused for arbitrary command execution.

A specially crafted image can be used to read or create/write arbitrary files on the host; possibly leading to arbitrary command execution.

Instance snapshots ignore the restricted.containers.lowlevel=block setting; allowing for arbitrary command execution on the Incus server by abusing lowlevel hooks such as raw.lxc and raw.qemu.

When providing invalid options to the wand option parser a small memory leak will occur.

When providing invalid options to the wand option parser a small memory leak will occur.

When providing invalid options to the wand option parser a small memory leak will occur.

When providing invalid options to the wand option parser a small memory leak will occur.

When providing invalid options to the wand option parser a small memory leak will occur.

When providing invalid options to the wand option parser a small memory leak will occur.

When providing invalid options to the wand option parser a small memory leak will occur.

When providing invalid options to the wand option parser a small memory leak will occur.

When providing invalid options to the wand option parser a small memory leak will occur.

When providing invalid options to the wand option parser a small memory leak will occur.

When providing invalid options to the wand option parser a small memory leak will occur.

When providing invalid options to the wand option parser a small memory leak will occur.

When providing invalid options to the wand option parser a small memory leak will occur.

When providing invalid options to the wand option parser a small memory leak will occur.

When providing invalid options to the wand option parser a small memory leak will occur.

When providing invalid options to the wand option parser a small memory leak will occur.

When providing invalid options to the wand option parser a small memory leak will occur.

When passing incorrect arguments in the distort operation a null pointer deference will occur.

When passing incorrect arguments in the distort operation a null pointer deference will occur.

When passing incorrect arguments in the distort operation a null pointer deference will occur.

When passing incorrect arguments in the distort operation a null pointer deference will occur.

When passing incorrect arguments in the distort operation a null pointer deference will occur.

When passing incorrect arguments in the distort operation a null pointer deference will occur.

When passing incorrect arguments in the distort operation a null pointer deference will occur.

When passing incorrect arguments in the distort operation a null pointer deference will occur.

When passing incorrect arguments in the distort operation a null pointer deference will occur.

When passing incorrect arguments in the distort operation a null pointer deference will occur.

When passing incorrect arguments in the distort operation a null pointer deference will occur.

When passing incorrect arguments in the distort operation a null pointer deference will occur.

When passing incorrect arguments in the distort operation a null pointer deference will occur.

When passing incorrect arguments in the distort operation a null pointer deference will occur.

When passing incorrect arguments in the distort operation a null pointer deference will occur.

When passing incorrect arguments in the distort operation a null pointer deference will occur.

When passing incorrect arguments in the distort operation a null pointer deference will occur.

When an allocation fails in CheckPrimitiveExtent this can result in a heap-use-after-free and result in a crash.

When an allocation fails in CheckPrimitiveExtent this can result in a heap-use-after-free and result in a crash.

When an allocation fails in CheckPrimitiveExtent this can result in a heap-use-after-free and result in a crash.

When an allocation fails in CheckPrimitiveExtent this can result in a heap-use-after-free and result in a crash.

When an allocation fails in CheckPrimitiveExtent this can result in a heap-use-after-free and result in a crash.

When an allocation fails in CheckPrimitiveExtent this can result in a heap-use-after-free and result in a crash.

When an allocation fails in CheckPrimitiveExtent this can result in a heap-use-after-free and result in a crash.

When an allocation fails in CheckPrimitiveExtent this can result in a heap-use-after-free and result in a crash.

When an allocation fails in CheckPrimitiveExtent this can result in a heap-use-after-free and result in a crash.

When an allocation fails in CheckPrimitiveExtent this can result in a heap-use-after-free and result in a crash.

When an allocation fails in CheckPrimitiveExtent this can result in a heap-use-after-free and result in a crash.

When an allocation fails in CheckPrimitiveExtent this can result in a heap-use-after-free and result in a crash.

When an allocation fails in CheckPrimitiveExtent this can result in a heap-use-after-free and result in a crash.

When an allocation fails in CheckPrimitiveExtent this can result in a heap-use-after-free and result in a crash.

When an allocation fails in CheckPrimitiveExtent this can result in a heap-use-after-free and result in a crash.

When an allocation fails in CheckPrimitiveExtent this can result in a heap-use-after-free and result in a crash.

When an allocation fails in CheckPrimitiveExtent this can result in a heap-use-after-free and result in a crash.

An crafted multi-frame can result in a heap buffer over-write when encoding it with the SF3 encoder.

An crafted multi-frame can result in a heap buffer over-write when encoding it with the SF3 encoder.

An crafted multi-frame can result in a heap buffer over-write when encoding it with the SF3 encoder.

An crafted multi-frame can result in a heap buffer over-write when encoding it with the SF3 encoder.

An crafted multi-frame can result in a heap buffer over-write when encoding it with the SF3 encoder.

An crafted multi-frame can result in a heap buffer over-write when encoding it with the SF3 encoder.

An crafted multi-frame can result in a heap buffer over-write when encoding it with the SF3 encoder.

An crafted multi-frame can result in a heap buffer over-write when encoding it with the SF3 encoder.

An crafted multi-frame can result in a heap buffer over-write when encoding it with the SF3 encoder.

An crafted multi-frame can result in a heap buffer over-write when encoding it with the SF3 encoder.

An crafted multi-frame can result in a heap buffer over-write when encoding it with the SF3 encoder.

An crafted multi-frame can result in a heap buffer over-write when encoding it with the SF3 encoder.

An crafted multi-frame can result in a heap buffer over-write when encoding it with the SF3 encoder.

An crafted multi-frame can result in a heap buffer over-write when encoding it with the SF3 encoder.

An crafted multi-frame can result in a heap buffer over-write when encoding it with the SF3 encoder.

An crafted multi-frame can result in a heap buffer over-write when encoding it with the SF3 encoder.

An crafted multi-frame can result in a heap buffer over-write when encoding it with the SF3 encoder.

Sending an excessively large header by an attacker could lead to a server-side DoS attack.

An authenticated client can crash the Hysteria server by advertising a very small QUIC max_datagram_frame_size and then triggering a UDP response from the server. When the server tries to send the UDP response back via QUIC DATAGRAM, quic-go returns DatagramTooLargeError. The server then attempts to fragment the Hysteria UDP message, but the fragmentation code does not handle the case where the UDP message header itself is larger than the maximum …

Hysteria's UDP relay treats the destination address as packet-scoped, but ACL and outbound policy are applied only once when a new UDP session is created. After an authenticated client opens a UDP session using an allowed first destination, later packets in the same Session ID can be sent to different destinations without re-running ACL evaluation. This allows an authenticated user to bypass server-side UDP ACL rules and reach localhost or …

The maintainer's recent fix in 6dd71e6a3c966867ef8c900d359a7df75789f410 (fix(subsonic): enforce playlist ownership on getPlaylist/deletePlaylist) added an ownership check based on playlist.UserID. However, playlist.UserID is derived from the first path segment of the attacker-controlled playlist ID, with no path containment on the resolved file path. Any authenticated Subsonic user can therefore bypass the ownership check and: Read any other user's playlist (name, comment, IsPublic flag, song list) by crafting a base64-encoded playlist ID …

A logic error in ServeCreateOrUpdatePlaylist allows any authenticated Subsonic user (including non-admin) to write playlist M3U content to an attacker-controlled absolute filesystem path on the gonic host, and to create intermediate directories with 0o777 permissions. The bug is independent of the playlist ownership IDOR fixed in 6dd71e6: it is an unreachable guard clause combined with no path containment in Store.Write.

The out_http output plugin allows the use of placeholders (such as ${tag}) in the endpoint configuration parameter. It was discovered that if the placeholder value is derived from untrusted user input, an attacker can maliciously control the destination hostname of the outbound HTTP requests made by Fluentd.

Fluentd allows dynamically constructing file paths using the ${tag} placeholder. It was discovered that validation for this placeholder was insufficient. If a Fluentd instance is configured to receive logs from untrusted sources and uses the ${tag} placeholder in file configurations (such as the path parameter in the out_file plugin), an attacker can inject path traversal characters (e.g., ../). When combined with certain formatting options, this vulnerability allows an attacker to …

Fluentd's Monitor Agent plugin (in_monitor_agent) exposes internal metrics and plugin information via a REST API. It was discovered that the API response (/api/plugins.json and related endpoints) unintentionally includes internal instance variables of loaded plugins. If any plugins store sensitive information—such as database passwords, API keys, or cloud credentials—in its instance variables, this information may be exposed in plain text to any user or system that has HTTP access to the …

Fluentd's in_http and in_forward plugins support receiving gzip-compressed data. While Fluentd correctly enforces size limits on the incoming compressed payloads (e.g., via body_size_limit or chunk_size_limit), it was discovered that there is no limit enforced on the size of the decompressed data. If a Fluentd instance is exposed to untrusted networks, an attacker can send a maliciously crafted, highly compressed payload. When Fluentd attempts to decompress this payload in memory, it …

The fluent-plugin-s3 plugin (specifically the in_s3 input plugin) supports reading and decompressing heavily compressed files (such as gzip, lzma2, and lzop) from Amazon S3. It was discovered that the plugin read the entire decompressed payload into memory at once without enforcing a strict size limit. If an attacker has sufficient permissions to upload files to the monitored S3 bucket, they can upload a maliciously crafted, highly compressed file. When Fluentd …

The fluent-plugin-opentelemetry plugin (specifically the in_opentelemetry HTTP input) lacked strict size limits on incoming requests. It was discovered that the plugin read the entire request body and decompressed payloads into memory without enforcing maximum size thresholds. If the OpenTelemetry ingestion endpoint is exposed to untrusted networks, an attacker can send an excessively large HTTP request or a maliciously crafted, highly compressed payload. When the plugin attempts to read or decompress …

The global policy read endpoint (GET /api/latest/fleet/policies/{policy_id}) performs authorization against an empty fleet.Policy{} struct with nil TeamID, then fetches any policy by ID from the database without verifying the fetched policy actually belongs to the global scope. This allows a user with observer-level access on any single team to read the full details of policies belonging to any other team, bypassing Fleet's team isolation model.

This vulnerability is an improper input neutralization issue leading to output manipulation, specifically, Terminal/ANSI Escape Sequence Injection and XML Injection: Terminal Output Spoofing: A malicious file whose name contains ANSI escape sequences can end up being included in flawfinder's standard terminal output, with many effects. For example, this might allow an attacker to hide critical scan results, falsely making it appear to a human reviewer that no security issues were …

The HTML and RSS output handlers in dosagelib/events.py write user-controlled content (comic text and page URLs) directly into generated files without proper HTML escaping. When a user scrapes a malicious webcomic and opens the generated HTML/RSS file, attacker-controlled JavaScript can execute in their browser. CWE: CWE-79 - Improper Neutralization of Input During Web Page Generation (Cross-site Scripting)

Prototype pollution in deepstream server v <=10.0.4. Potential privilege escalation from any authenticated user with write permission to any record.

The Rust Security Response Team was notified that Cargo incorrectly handled symlinks inside of crate tarballs downloaded from third-party registries, allowing a malicious crate to override the source code of another crate from the same registry. This vulnerability is tracked as CVE-2026-5223. The severity of the vulnerability is medium for users of third-party registries. Users of crates.io are not affected, as crates.io forbids uploading crates containing any symlink.

The Rust Security Response Team was notified that Cargo incorrectly normalized the URLs of third-party registries using the [sparse index protocol][1]. If a hosting provider allowed multiple registries to be hosted with arbitrary names within the same domain, an attacker able to publish crates in a registry could obtain the credentials of others users of the same registry. This vulnerability is tracked as CVE-2026-5222. The severity of the vulnerability is …

View::_getElementFileName() does not check that the resolved element path is within the application/plugin view template paths. When element names are created with specifically crafted user-supplied data this weakness can be leveraged to include other PHP files on the server.

Blnk API key endpoints had an authorization issue that allowed non-master API keys to perform key-management actions outside their intended authorization boundary. In affected versions, API key operations trusted caller-controlled request values for owner and scope decisions. As a result, a non-master API key could potentially manage keys for another owner by supplying a different owner value, or create a more privileged API key by requesting broader scopes than it …

A directory traversal vulnerability exists in the production static file server of better-helperjs (<= 3.0.5). Attackers can read arbitrary files located in adjacent directory structures that share the same string prefix as the intended static root directory.

In backpropagate >= 1.1.0, the optional Reflex web UI (pip install backpropagate[ui], launched via backprop ui) exposes a training control plane: dataset upload, model load, training start/stop, multi-run orchestration, GGUF export, and HuggingFace Hub push. The CLI accepts two operator-facing flags intended as security controls: –auth user:pass — documented as "require HTTP Basic authentication on every request to the UI." –share — documented as "expose the UI on a public …

In backpropagate >= 1.1.0, the optional Reflex web UI (pip install backpropagate[ui], launched via backprop ui) exposes a training control plane: dataset upload, model load, training start/stop, multi-run orchestration, GGUF export, and HuggingFace Hub push. The CLI accepts two operator-facing flags intended as security controls: –auth user:pass — documented as "require HTTP Basic authentication on every request to the UI." –share — documented as "expose the UI on a public …

Due to lack of canonicalization of domains in very specific edge cases an access control rule may be skipped when it should match a request.

The limit container paths directive in apptainer.conf is intended to allow a system administrator limit the paths from which containers can be run, under setuid mode. Due to incorrect matching of a path string, sibling directories with similar names may incorrectly be allowed. For example, the configuration: limit container paths = /data/safe Will also allow containers in /data/safe-but-unsafe to be run.

The administrative proxy route (cmsproxy) in Aimeos Pagible CMS is vulnerable to a Server-Side Request Forgery (SSRF) attack via DNS Rebinding. A Time-of-Check to Time-of-Use (TOCTOU) race condition exists between the URL validation phase and the actual HTTP request phase, allowing attackers to access internal network resources and cloud metadata endpoints.

The preAuthEncoding function in @sigstore/core uses Node.js 'ascii' encoding when converting the PAE (Pre-Authentication Encoding) string to bytes. This allows payloadType to be mutated after signing without invalidating the signature, breaking the type-binding guarantee that DSSE is designed to provide. In packages/core/src/dsse.ts, the PAE function builds a string containing payloadType and then encodes it with Buffer.from(prefix, 'ascii'). In Node.js, 'ascii' encoding for string-to-Buffer is equivalent to 'latin1', which truncates characters …

@microsoft/kiota-http-fetchlibrary's RedirectHandler is documented as stripping Authorization and Cookie from cross-origin redirect targets, but the default scrubSensitiveHeaders callback in RedirectHandlerOptions uses case-sensitive property deletion (delete headers.Authorization, delete headers.Cookie) on a headers object that FetchRequestAdapter.getRequestFromRequestInformation has already lower-cased. The delete therefore targets keys that do not exist, the scrub is a no-op, and any Bearer token or Cookie attached by a kiota-generated SDK is forwarded to an attacker-controlled host across a …

A command injection vulnerability existed in the Maven scanning flow of cdxgen before version 12.4.3. When cdxgen scanned an attacker-controlled Maven project, repository-controlled paths could be used in the Maven command construction. In affected versions, some Maven invocations were executed with shell: true. A directory name containing shell metacharacters could therefore be interpreted by the shell instead of being treated only as a filesystem path. This could allow an attacker …

@cardano402/mcp-server versions <= 0.1.1 ship three security gaps that can lead to unauthorized fund movement when the package is used as designed (an MCP server exposing Cardano payment tools to an

The Package.Unmarshal() function in pkg/types/alpine/apk.go decompresses the signature and control gzip members of an APK file into in-memory buffers without bounding the total decompressed size. The existing max_apk_metadata_size check (default 1MB) is only applied to individual tar entry header sizes after decompression completes, so it does not prevent a decompression bomb from consuming unbounded heap memory. An attacker can craft a gzip stream that compresses at a ~1000:1 ratio (e.g., …

BaggagePropagator::extract_with_context in opentelemetry_sdk did not enforce the W3C Baggage size limits before parsing an inbound baggage header. A large attacker-controlled header could cause unnecessary CPU work and short-lived heap allocations while parsing entries that would later be discarded by the SDK's baggage storage limits. The SDK now applies limits aligned with the W3C Baggage limits: 64 list-members 8192 bytes total

Description An Improper Verification of Cryptographic Signature (CWE-347) issue in OpenAM's RADIUS authentication module allows an unauthenticated network attacker to spoof an Access-Accept response and obtain an OpenAM session for any RADIUS username, without knowing the configured shared secret. This affects OpenAM Community Edition through version 16.0.6 and was patched in version 16.1.1. The RADIUS client opens an unconnected datagram socket and treats the first UDP datagram delivered to its …

Description A Deserialization of Untrusted Data (CWE-502) issue exists in OpenAM's Push Notification SNS callback resource. The REST route that handles SNS push messages is mounted with anonymous access and, when a supplied message identifier has expired from the in-memory dispatcher, falls back to a CTS-stored predicate blob whose top-level keys are treated as Java class names and passed to Class.forName(…) before attacker-controlled JSON is deserialized via Jackson. This impacts …

Description An Authorization Bypass Through User-Controlled Key (CWE-639) exists in OpenAM's stateful OAuth2 token-read path. Under certain conditions, this may allow an attacker to forge OAuth2 bearer tokens and OIDC ID tokens with arbitrary subject, client, realm, and scope. This affects OpenAM Community Edition through version 16.0.6. The OAuth2 token-read path reads caller-supplied token identifiers from the shared Core Token Store (CTS) without placing them in an OAuth-only namespace and …

nextflow auth login persists Seqera Platform OIDC tokens to ${NXF_HOME:-~/.nextflow}/seqera-auth.config. The file is created via Java NIO without specifying file permissions, so under the default umask 022 it lands at mode 0644 (world-readable). On a multi-user POSIX host — typically an HPC login node, shared workstation, or jump host — any local user able to traverse the victim's home directory can read the file and obtain a valid Platform bearer …

The /list_relationships and /retrieve_graph_neighborhood endpoints call getAuthenticatedUserId (confirming a valid session exists) but do not pass the resolved user ID into the Supabase query as an .eq("user_id", userId) filter. As a result, queries return rows from all users rather than scoping to the authenticated caller's data.

UnsafeBlitFormatterBase<T>.Deserialize reads an attacker-controlled byteLength from an extension payload and allocates an array based on that value before validating it against the extension header length or remaining payload bytes. The outer extension header is bounded by available input, but that bound is not used to constrain the inner byteLength before allocation. A very small payload can therefore request a very large T[] allocation.

MessagePack-CSharp's typeless deserialization includes MessagePackSerializerOptions.ThrowIfDeserializingTypeIsDisallowed(Type) as a safety check for dangerous types. The default implementation checks the outer type name, but it does not recursively inspect array element types or generic type arguments. As a result, a type that would be blocked directly can be wrapped inside an array or constructed generic type and pass the outer type check. The formatter machinery can then materialize formatters for the inner blocked …

MessagePack-CSharp's multi-dimensional array formatters read dimension lengths directly from the payload and allocate T[,], T[,,], or T[,,,] before validating that the dimension product matches the encoded element count. The formatter reads a guarded element array header, but allocation of the target multi-dimensional array happens before the dimensions are checked against that element count. A small payload can therefore declare large dimensions, provide an empty or tiny inner array, and cause …

MessagePackReader.TrySkip() recursively descends into nested arrays and maps without incrementing the reader depth or calling the configured depth checks. This bypasses MessagePackSecurity.MaximumObjectGraphDepth, the library's documented protection against deeply nested object graphs. Many generated and dynamic formatters call reader.Skip() when they encounter unknown map keys, unknown array members, ignored fields, or data that should be skipped for forward compatibility. A deeply nested value in one of these skipped positions can therefore …

When MessagePack-CSharp decompresses Lz4Block or Lz4BlockArray payloads, it reads declared uncompressed lengths from the wire and allocates output buffers based on those lengths before validating that the compressed data is valid or that the declared expansion is reasonable. A small payload can claim a very large uncompressed length and force a large allocation before LZ4 decoding begins.

MessagePack-CSharp's JSON conversion helpers contain multiple recursion paths that do not consistently enforce a depth limit. These paths are in the JSON conversion component rather than normal typed MessagePack deserialization. Three related issues are covered by this advisory: MessagePackSerializer.ConvertFromJson recursively processes nested JSON arrays and objects in FromJsonCore() without consulting MessagePackSecurity.MaximumObjectGraphDepth. TinyJsonReader.ReadNextToken() recursively consumes comma and colon separator characters, allowing even malformed JSON with long separator runs to consume one …

InterfaceLookupFormatter<TKey,TElement> constructs an internal Dictionary<TKey, IGrouping<TKey,TElement>> with the default equality comparer instead of the security-aware comparer supplied by options.Security.GetEqualityComparer<TKey>(). Other hash-based collection formatters use the security-aware comparer when MessagePackSecurity.UntrustedData is configured. This formatter omission allows hash-collision CPU denial of service against ILookup<TKey,TElement> even when the application has opted into the untrusted-data security posture.

ExpandoObjectFormatter.Deserialize populates System.Dynamic.ExpandoObject by calling IDictionary<string, object>.Add for each map entry. ExpandoObject internally maintains member names in array-like structures, so inserting many distinct keys can require repeated linear scans and array copies. For large attacker-controlled maps, this produces quadratic CPU and allocation behavior. The issue is especially surprising because ExpandoObjectResolver.Options is configured with MessagePackSecurity.UntrustedData, but collision-resistant dictionary comparers cannot protect ExpandoObject insertion internals.

Runtime-generated union deserializers emitted by DynamicUnionResolver do not call MessagePackSecurity.DepthStep(ref reader) and do not decrement reader.Depth around recursive deserialization and skip paths. This means union deserialization does not consistently participate in the maximum object graph depth enforcement that protects other recursive formatter paths. For unknown union keys, the emitted deserializer calls reader.Skip() on attacker-controlled data without an enclosing depth step.

MessagePackReader.ReadDateTime() can allocate stack memory based on an attacker-controlled MessagePack extension length. In the slow path for timestamp extension parsing, the computed tokenSize includes the extension body length from the wire and is used in a stackalloc operation before the extension length is validated as one of the valid timestamp sizes. A very small payload can claim a large timestamp extension body and cause a stack allocation large enough to …

The parameterless MessagePackInputFormatter() constructor uses default serializer options, which resolve to MessagePackSerializerOptions.Standard with MessagePackSecurity.TrustedData. The formatter is designed for ASP.NET Core MVC request bodies, which commonly cross an HTTP trust boundary. This insecure default can expose applications to denial-of-service attacks that MessagePackSecurity.UntrustedData is intended to mitigate, such as hash-collision attacks against dictionary-like model properties.

When verifying an uploaded certificate, lemur/certificates/verify.py extracts the CRL Distribution Point URL and the OCSP responder URL directly from the certificate's extensions and issues outbound requests to those URLs without scheme restriction or destination allow-listing. An authenticated user holding the operator role (required by StrictRolePermission on POST /certificates/upload) can craft a certificate whose extensions point at internal services - instance metadata endpoints, internal Kubernetes API servers, RFC1918 hosts, link-local addresses …

Field | Value – | – Title | Lemur 1.9.0: any SSO-authenticated user achieves AWS IAM compromise and permanent PKI key access via ACME acme_url SSRF and creator-equality IDOR Component | lemur/lemur/plugins/lemur_acme/acme_handlers.py:161-201 (SSRF), lemur/lemur/certificates/views.py:734 (IDOR), lemur/lemur/auth/views.py:300-308 (SSO auto-provision) CWE | CWE-918 (SSRF) + CWE-639 (Authorization Bypass Through User-Controlled Key) + CWE-285 (Improper Authorization) Attack Prerequisite | A valid SSO session against the deployment's IdP. Lemur auto-provisions any new SSO identity …

lemur.users.service.update() writes a user's new password as plaintext to the users.password column. The User model wires bcrypt hashing to SQLAlchemy's before_insert event but registers no equivalent listener for before_update, and service.update() does not call user.hash_password() after assigning the new value. Every password change performed through the admin-gated PUT /api/1/users/<id> endpoint persists the user's password to the database in cleartext.

The PUT /api/1/roles/<id> handler in lemur/roles/views.py gates only on RoleMemberPermission(role_id).can(), which is satisfied for any user who is already a member of the target role. The handler then passes data["users"] and data["name"] directly to service.update(), permitting any role member to rewrite that role's membership list and name. The companion DELETE handler on the same resource is correctly gated by @admin_permission.require; the asymmetry between PUT and DELETE on identical resources indicates …

StrictRolePermission and AuthorityCreatorPermission in lemur/auth/permissions.py call flask_principal.Permission.init() with zero Needs when their config flags are unset. Both flags defaulted to False in code prior to the fix, so this was the state of any Lemur install that hadn't explicitly opted in. Flask-Principal's Permission.allows() returns True whenever self.needs is empty. The .can() gate therefore passes for every authenticated identity, including the lowest-privilege role Lemur ships (read-only). A user holding only read-only …

langgraph-sdk constructs HTTP request paths for resource operations by interpolating caller-supplied identifier values into URL templates. Without sanitization of those values, identifiers that contain characters with special meaning in URL paths could cause the resulting request to address a different resource (and potentially a different resource type) than the SDK method's call site indicates. In deployments where the SDK receives identifier values that originate from untrusted sources, this could result …

LangGraph's JsonPlusSerializer can reconstruct Python objects from JSON checkpoint payloads. Under conditions where someone could modify checkpoint bytes at rest in the backing store, the deserialization path could reconstruct objects beyond what the application expects, which could in turn result in code execution at checkpoint load time. This is a defense-in-depth issue. The affected behavior is reachable only when checkpoint bytes at rest in the backing store can be modified …

In justhtml 0.9.0 through 1.21.0, to_markdown() renders <code> text (and <pre> text inside a link) as an inline Markdown code span whose only protection is backtick-fence length. A blank line (\n\n) in that text terminates the inline span in any compliant Markdown renderer, so attacker-controlled text that survived HTML sanitization is emitted unescaped after the blank line and is re-parsed as live raw HTML/Markdown — yielding XSS in the default …

A missing check in the DCM decoder could result in an image with invalid dimensions and that could cause crashes in other operations.

A missing check in the DCM decoder could result in an image with invalid dimensions and that could cause crashes in other operations.

A missing check in the DCM decoder could result in an image with invalid dimensions and that could cause crashes in other operations.

A missing check in the DCM decoder could result in an image with invalid dimensions and that could cause crashes in other operations.

A missing check in the DCM decoder could result in an image with invalid dimensions and that could cause crashes in other operations.

A missing check in the DCM decoder could result in an image with invalid dimensions and that could cause crashes in other operations.

A missing check in the DCM decoder could result in an image with invalid dimensions and that could cause crashes in other operations.

A missing check in the DCM decoder could result in an image with invalid dimensions and that could cause crashes in other operations.

A missing check in the DCM decoder could result in an image with invalid dimensions and that could cause crashes in other operations.

A missing check in the DCM decoder could result in an image with invalid dimensions and that could cause crashes in other operations.

A missing check in the DCM decoder could result in an image with invalid dimensions and that could cause crashes in other operations.

A missing check in the DCM decoder could result in an image with invalid dimensions and that could cause crashes in other operations.

A missing check in the DCM decoder could result in an image with invalid dimensions and that could cause crashes in other operations.

A missing check in the DCM decoder could result in an image with invalid dimensions and that could cause crashes in other operations.

A missing check in the DCM decoder could result in an image with invalid dimensions and that could cause crashes in other operations.

A missing check in the DCM decoder could result in an image with invalid dimensions and that could cause crashes in other operations.

A missing check in the DCM decoder could result in an image with invalid dimensions and that could cause crashes in other operations.

A missing check for maximum memory request in AcquireAlignedMemory could trigger an out-of-Memory condition.

A missing check for maximum memory request in AcquireAlignedMemory could trigger an out-of-Memory condition.

A missing check for maximum memory request in AcquireAlignedMemory could trigger an out-of-Memory condition.

A missing check for maximum memory request in AcquireAlignedMemory could trigger an out-of-Memory condition.

A missing check for maximum memory request in AcquireAlignedMemory could trigger an out-of-Memory condition.

A missing check for maximum memory request in AcquireAlignedMemory could trigger an out-of-Memory condition.

A missing check for maximum memory request in AcquireAlignedMemory could trigger an out-of-Memory condition.

A missing check for maximum memory request in AcquireAlignedMemory could trigger an out-of-Memory condition.

A missing check for maximum memory request in AcquireAlignedMemory could trigger an out-of-Memory condition.

A missing check for maximum memory request in AcquireAlignedMemory could trigger an out-of-Memory condition.

A missing check for maximum memory request in AcquireAlignedMemory could trigger an out-of-Memory condition.

A missing check for maximum memory request in AcquireAlignedMemory could trigger an out-of-Memory condition.

A missing check for maximum memory request in AcquireAlignedMemory could trigger an out-of-Memory condition.

A missing check for maximum memory request in AcquireAlignedMemory could trigger an out-of-Memory condition.