DEV Community: CVE Reports

CVE-2023-49316: CVE-2023-49316: Denial of Service via Unbounded Degree in phpseclib Binary Finite Fields

CVE Reports — Sat, 09 May 2026 06:10:29 +0000

CVE-2023-49316: Denial of Service via Unbounded Degree in phpseclib Binary Finite Fields

Vulnerability ID: CVE-2023-49316
CVSS Score: 7.5
Published: 2026-05-08

The phpseclib cryptographic library version 3.x prior to 3.0.34 contains a Denial of Service (DoS) vulnerability in its mathematical field generation logic. When parsing maliciously crafted X.509 certificates or PKCS#8 private keys specifying Elliptic Curve parameters over a binary finite field, the library fails to validate the degree parameter. This flaw allows a remote attacker to force the PHP application to perform unbounded memory allocations, exhausting server resources and terminating the application worker process.

TL;DR

phpseclib 3.x before 3.0.34 fails to bound the degree parameter when parsing Elliptic Curve keys over binary fields, allowing unauthenticated attackers to cause a fatal Out-Of-Memory (OOM) crash via a crafted ASN.1 payload.

⚠️ Exploit Status: POC

Technical Details

CWE ID: CWE-834
Attack Vector: Network
CVSS v3.1: 7.5
EPSS Score: 0.15%
Impact: High (Availability)
Exploit Status: Proof-of-Concept
CISA KEV: Not Listed

Affected Systems

phpseclib 3.0.x
phpseclib: >= 3.0.0, < 3.0.34 (Fixed in: 3.0.34)

Code Analysis

Commit: 964d781

Add degree bounds checking to BinaryField constructor to prevent OOM denial of service

Exploit Details

Research Paper: X.509DoS: Exploiting and Detecting Denial-of-Service Vulnerabilities in X.509 Certificate Parsing

Mitigation Strategies

Upgrade phpseclib to version 3.0.34 or later.
Implement application-level exception handling for OutOfBoundsException during cryptographic material parsing.
Enforce strict file size limits on endpoints accepting user-uploaded certificate files.

Remediation Steps:

Identify the project dependencies using composer: composer show phpseclib/phpseclib.
Update the composer requirement to pull the latest 3.0.x patch: composer require phpseclib/phpseclib:^3.0.34.
Execute composer update.
Deploy the updated composer.lock file to production environments.
Audit all calls to EC::loadFormat() to ensure proper try...catch blocks encompass the parsing logic.

References

Read the full report for CVE-2023-49316 on our website for more details including interactive diagrams and full exploit analysis.

GHSA-MV93-W799-CJ2W: GHSA-MV93-W799-CJ2W: Remote Code Execution via Config Section Injection in GitPython

CVE Reports — Sat, 09 May 2026 05:40:29 +0000

GHSA-MV93-W799-CJ2W: Remote Code Execution via Config Section Injection in GitPython

Vulnerability ID: GHSA-MV93-W799-CJ2W
CVSS Score: 7.8
Published: 2026-05-08

GitPython versions prior to 3.1.50 are vulnerable to a newline injection attack in the config_writer() and set_value() methods. An incomplete fix for CVE-2026-44244 failed to sanitize the configuration section parameter, allowing an attacker to inject malicious Git configuration blocks such as [core] and override the hooksPath. This leads to unauthenticated remote code execution when subsequent Git operations trigger the injected hooks.

TL;DR

Newline injection in GitPython's config_writer section parameter allows attackers to override core.hooksPath and achieve Remote Code Execution.

⚠️ Exploit Status: POC

Technical Details

CWE ID: CWE-94, CWE-150
Attack Vector: Local (via Application Input)
CVSS Base Score: 7.8 (High)
Exploit Status: Proof of Concept Available
Impact: Remote Code Execution via Git Hooks
Patched Version: 3.1.50

Affected Systems

GitPython pip package
GitPython: < 3.1.50 (Fixed in: 3.1.50)

Mitigation Strategies

Upgrade GitPython pip package to version 3.1.50 or later.
Implement application-level input validation to sanitize user input passed to GitPython API.
Reject or strip newline characters (\n, \r) from configuration section variables.
Monitor .git/config files for anomalous structures and unexpected hooksPath entries.

Remediation Steps:

Identify all projects utilizing the GitPython library within the environment.
Update the GitPython dependency to >= 3.1.50 via package manager (e.g., pip install --upgrade GitPython).
Review application source code for calls to config_writer().set_value() and config_writer().add_section().
Ensure input passed to these methods is heavily sanitized if derived from external sources.
Deploy file integrity monitoring rules to alert on modifications to repository .git/config files introducing new hooksPath directives.

References

Read the full report for GHSA-MV93-W799-CJ2W on our website for more details including interactive diagrams and full exploit analysis.

CVE-2026-6860: CVE-2026-6860: Unbounded SNI Cache Growth in Eclipse Vert.x

CVE Reports — Sat, 09 May 2026 05:10:28 +0000

CVE-2026-6860: Unbounded SNI Cache Growth in Eclipse Vert.x

Vulnerability ID: CVE-2026-6860
CVSS Score: 5.3
Published: 2026-05-09

Eclipse Vert.x suffers from an uncontrolled resource consumption vulnerability within its Server Name Indication (SNI) processing logic. When server-side SNI is enabled alongside a wildcard TLS certificate, unauthenticated remote attackers can exhaust server memory by initiating handshakes with continuous unique hostname values, ultimately resulting in a Denial of Service (DoS).

TL;DR

A flaw in the SNI caching mechanism of Eclipse Vert.x allows remote attackers to trigger out-of-memory (OOM) conditions. By sending numerous TLS ClientHello messages with uniquely generated hostnames matching a wildcard certificate, an attacker bypasses cache hits and forces the unbounded allocation of SslContext objects in JVM memory.

⚠️ Exploit Status: POC

Technical Details

CWE ID: CWE-400
Attack Vector: Network
CVSS v4.0 Score: 6.9 (Medium)
CVSS v3.1 Score: 5.3 (Medium)
Impact: Denial of Service (OOM)
Exploit Status: Proof of Concept
EPSS Percentile: 6.48%
KEV Status: Not Listed

Affected Systems

Eclipse Vert.x Applications
Java Virtual Machine (JVM) instances running Vert.x TLS
Microservices utilizing io.vertx:vertx-core for SNI termination
io.vertx:vertx-core: >= 4.3.4, <= 4.3.8 (Fixed in: 4.3.9)
io.vertx:vertx-core: >= 4.4.0, <= 4.4.9 (Fixed in: 4.4.10)
io.vertx:vertx-core: >= 4.5.0, <= 4.5.26 (Fixed in: 4.5.27)
io.vertx:vertx-core: >= 5.0.0, <= 5.0.11 (Fixed in: 5.0.12)

Mitigation Strategies

Upgrade the io.vertx:vertx-core library to a patched version.
Disable server-side SNI via configuration if not strictly required.
Offload TLS termination to a network edge component (e.g., Nginx, HAProxy, AWS ALB).
Avoid the use of wildcard certificates where strict domain matching is possible.
Implement rate limiting for new TLS handshake connections per source IP address.

Remediation Steps:

Identify all projects and microservices depending on io.vertx:vertx-core.
Update dependency management files (pom.xml for Maven, build.gradle for Gradle) to specify the appropriate patched version.
Recompile and execute integration test suites to verify compatibility with the new minor release.
Deploy the updated application artifact to the staging environment and monitor heap usage.
Promote the patched application to the production environment.

References

Read the full report for CVE-2026-6860 on our website for more details including interactive diagrams and full exploit analysis.

GHSA-V6WJ-C83F-V46X: GHSA-v6wj-c83f-v46x: Critical OS Command Injection in @profullstack/mcp-server domain_lookup Module

CVE Reports — Sat, 09 May 2026 04:40:29 +0000

GHSA-v6wj-c83f-v46x: Critical OS Command Injection in @profullstack/mcp-server domain_lookup Module

Vulnerability ID: GHSA-V6WJ-C83F-V46X
CVSS Score: 9.8
Published: 2026-05-09

A critical unauthenticated OS Command Injection vulnerability (CWE-78) exists in the @profullstack/mcp-server npm package, specifically within the domain_lookup module. The vulnerability allows remote attackers to execute arbitrary commands on the host system via crafted HTTP requests.

TL;DR

The @profullstack/mcp-server package (versions <= 1.4.12) is vulnerable to unauthenticated OS Command Injection. The domain_lookup module unsafely concatenates user-supplied input into a shell command, enabling remote code execution.

⚠️ Exploit Status: POC

Technical Details

CWE ID: CWE-78
Attack Vector: Network
CVSS Score: 9.8 (Critical)
Impact: Arbitrary Remote Code Execution
Exploit Status: Proof of Concept Available
Privileges Required: None

Affected Systems

@profullstack/mcp-server npm package
@profullstack/mcp-server: <= 1.4.12

Mitigation Strategies

Replace child_process.exec with child_process.spawn to pass arguments safely as arrays.
Implement strict regex-based input validation for domain names enforcing RFC 1035 compliance.
Enforce global authentication middleware across all modular endpoints.
Execute the Node.js server process under a dedicated, unprivileged user account.

Remediation Steps:

Identify all deployments running @profullstack/mcp-server versions <= 1.4.12.
Modify service.js to utilize child_process.spawn instead of execAsync.
Implement a regex validation check (e.g., /^[a-zA-Z0-9.-]+$/) on the domains and keywords input arrays.
Restart the Node.js application process.
Review system logs and process trees for unauthorized executions or shell spawns indicating prior compromise.

References

Read the full report for GHSA-V6WJ-C83F-V46X on our website for more details including interactive diagrams and full exploit analysis.

GHSA-QHH4-458H-XWH2: GHSA-qhh4-458h-xwh2: Credential Leakage via Origin Validation Error in cdxgen

CVE Reports — Fri, 08 May 2026 22:10:29 +0000

GHSA-qhh4-458h-xwh2: Credential Leakage via Origin Validation Error in cdxgen

Vulnerability ID: GHSA-QHH4-458H-XWH2
CVSS Score: 5.3
Published: 2026-05-08

The @cyclonedx/cdxgen package is vulnerable to credential leakage due to improper Docker registry origin validation. A flaw in how registry authentication endpoints are matched against configured credentials allows arbitrary downstream registries to capture private credentials.

TL;DR

Versions 9.9.5 through 12.3.2 of @cyclonedx/cdxgen leak Docker registry credentials due to an insecure substring matching implementation. Upgrading to version 12.3.3 resolves the vulnerability by introducing strict hostname normalization.

⚠️ Exploit Status: POC

Technical Details

CWE ID: CWE-346, CWE-522
Attack Vector: Network
CVSS v4.0: 5.3 (Medium)
Impact: Credential Leakage
Exploit Status: Proof of Concept
KEV Status: Not Listed

Affected Systems

macOS / Linux / Windows running Node.js and @cyclonedx/cdxgen < 12.3.3
CI/CD pipelines utilizing vulnerable versions for SBOM generation
Docker Daemon instances invoked by vulnerable cdxgen processes
@cyclonedx/cdxgen: >= 9.9.5, < 12.3.3 (Fixed in: 12.3.3)

Code Analysis

Commit: b1e1798

Fix: Implement strict hostname normalization and validation for Docker registry authentication mapping.

Mitigation Strategies

Upgrade @cyclonedx/cdxgen to version 12.3.3 or later.
Rotate Docker registry credentials used in environments that previously ran vulnerable versions.
Implement strict egress network filtering on CI/CD build nodes to prevent unauthorized outbound connections.
Utilize dynamic, short-lived authentication tokens instead of static configurations in ~/.docker/config.json.

Remediation Steps:

Identify all global and local installations of @cyclonedx/cdxgen across developer machines and build servers.
Execute the update command: npm install -g @cyclonedx/cdxgen@latest
Verify the installed version using: cdxgen --version (Must output >= 12.3.3).
Revoke and regenerate service account tokens used by CI/CD pipelines for internal Docker registries.

References

Read the full report for GHSA-QHH4-458H-XWH2 on our website for more details including interactive diagrams and full exploit analysis.

CVE-2026-32689: CVE-2026-32689: Denial of Service in Phoenix Framework LongPoll Transport via NDJSON Payload Amplification

CVE Reports — Fri, 08 May 2026 20:40:29 +0000

CVE-2026-32689: Denial of Service in Phoenix Framework LongPoll Transport via NDJSON Payload Amplification

Vulnerability ID: CVE-2026-32689
CVSS Score: 8.7
Published: 2026-05-08

The Phoenix Framework contains a high-severity Denial of Service vulnerability in its LongPoll transport mechanism. The vulnerability is caused by unbounded memory allocation when processing Newline Delimited JSON (NDJSON) payloads. Unauthenticated attackers can trigger Out-Of-Memory conditions on the host BEAM node, terminating all active sessions by forcing the server to evaluate excessive newline characters.

TL;DR

Unauthenticated remote attackers can crash Phoenix Framework nodes by sending an 8MB NDJSON payload consisting entirely of newline characters to the LongPoll endpoint, triggering memory exhaustion via eager string evaluation.

⚠️ Exploit Status: POC

Technical Details

CWE ID: CWE-770
Attack Vector: Network
CVSS Base Score: 8.7 (High)
EPSS Score: 0.00045
Impact: Denial of Service (Node Crash)
Exploit Status: Unexploited / PoC Only
CISA KEV: No

Affected Systems

Phoenix Framework (phoenix Hex package) < 1.7.22
Phoenix Framework (phoenix Hex package) 1.8.x < 1.8.6
Any Elixir application exposing Phoenix LiveView endpoints via LongPoll
Phoenix Framework: >= 1.7.0, < 1.7.22 (Fixed in: 1.7.22)
Phoenix Framework: >= 1.8.0, < 1.8.6 (Fixed in: 1.8.6)

Code Analysis

Commit: 1a67c61

Server-side fix to replace String.split with String.splitter for lazy parsing.

Replaces String.split/2 and Enum.map with String.splitter/2 and Enum.find

Commit: 912ea18

Client-side fix to implement MAX_LONGPOLL_BATCH_SIZE in JavaScript.

Adds logic to split large longpoll message batches into chunks of 100

Mitigation Strategies

Upgrade Phoenix framework to patched versions (1.7.22 or 1.8.6).
Disable the LongPoll transport in Phoenix.Socket configurations if WebSocket connections are sufficient.
Implement network-level size restrictions or WAF rules to drop anomalous POST requests containing excessive application/x-ndjson payloads.

Remediation Steps:

Identify the current version of the phoenix Hex package running in your application environments.
Update the mix.exs dependencies to require at least ~> 1.7.22 or ~> 1.8.6.
Run mix deps.get and mix deps.compile to fetch and compile the updated framework code.
If patching is delayed, modify the endpoint module to set longpoll: false inside the socket declarations.

References

Read the full report for CVE-2026-32689 on our website for more details including interactive diagrams and full exploit analysis.

CVE-2026-44499: CVE-2026-44499: Permanent Block Discovery Halt in Zebra via Gossip Queue Saturation

CVE Reports — Fri, 08 May 2026 20:10:29 +0000

CVE-2026-44499: Permanent Block Discovery Halt in Zebra via Gossip Queue Saturation

Vulnerability ID: CVE-2026-44499
CVSS Score: 8.7
Published: 2026-05-08

CVE-2026-44499 is a composite Denial of Service (DoS) vulnerability affecting Zebra, the Rust implementation of a Zcash full node. By exploiting architectural flaws in the peer-to-peer (P2P) communication stack, an unauthenticated attacker can saturate internal message queues and poison the chain discovery process, permanently isolating the target node from the network.

TL;DR

Unauthenticated attackers can permanently halt block discovery in Zebra nodes prior to v4.4.0 by saturating the P2P gossip queue and providing unpenalized empty responses to synchronization requests.

Technical Details

CVSS Score: 8.7
CWE ID: CWE-770
Attack Vector: Network
Exploit Status: None
KEV Status: Not Listed
Authentication: None Required

Affected Systems

Zebra < 4.4.0
Zebra: < 4.4.0 (Fixed in: 4.4.0)

Code Analysis

Commit: 602c8b7

Implementation of Stall Tracking and peer penalization logic

Commit: 85303ea

Allocation hardening and size bounds enforcement

Mitigation Strategies

Upgrade to Zebra version 4.4.0 or later.
Implement network-level rate limiting for inbound P2P connections.
Monitor node synchronization metrics for abrupt halts in block height progression.

Remediation Steps:

Stop the affected Zebra service gracefully.
Update the Zebra binary to version 4.4.0 via your package manager or by compiling from the official repository.
Restart the Zebra service and monitor the logs to verify successful synchronization with the network.

References

Read the full report for CVE-2026-44499 on our website for more details including interactive diagrams and full exploit analysis.

CVE-2026-6322: CVE-2026-6322: Host Confusion via Interpretation Conflict in fast-uri

CVE Reports — Fri, 08 May 2026 19:40:29 +0000

CVE-2026-6322: Host Confusion via Interpretation Conflict in fast-uri

Vulnerability ID: CVE-2026-6322
CVSS Score: 7.5
Published: 2026-05-08

The fast-uri library exhibits an interpretation conflict vulnerability due to improper handling of percent-encoded authority delimiters during normalization. This flaw enables attackers to bypass domain validation and perform host confusion attacks against downstream components.

TL;DR

fast-uri <= 3.1.1 improperly decodes percent-encoded delimiters (like %40) in the host component without re-encoding them, causing downstream parsers to misinterpret the URI structure. Upgrading to 3.1.2 resolves the issue.

⚠️ Exploit Status: POC

Technical Details

CWE ID: CWE-436
Attack Vector: Network
CVSS v3.1: 7.5 (High)
EPSS Score: 0.00029
Impact: Integrity Subversion / SSRF Bypass
Exploit Status: Proof of Concept
CISA KEV: Not Listed

Affected Systems

fast-uri
fast-uri: <= 3.1.1 (Fixed in: 3.1.2)

Code Analysis

Commit: 6c86c17

Re-escape gen-delims in host

Commit: 919dd8e

Version 3.1.2 Bump

Exploit Details

GitHub Advisory: Technical root cause and reproduction steps mapping the SSRF bypass payload.

Mitigation Strategies

Upgrade fast-uri to version 3.1.2 or later.
Implement robust pre-validation checks rejecting URIs with percent-encoded reserved characters in the host.
Unify URI parsing logic to use the exact same library for both security validation and request execution.

Remediation Steps:

Identify all projects depending on fast-uri via package-lock.json or yarn.lock.
Execute the package manager update command targeting fast-uri@3.1.2.
Run regression tests on URI parsing and normalization workflows.
Deploy the updated application build to production environments.

References

Read the full report for CVE-2026-6322 on our website for more details including interactive diagrams and full exploit analysis.

CVE-2026-43944: CVE-2026-43944: Arbitrary Local Code Execution in electerm via Malicious Deep Links

CVE Reports — Fri, 08 May 2026 19:10:30 +0000

CVE-2026-43944: Arbitrary Local Code Execution in electerm via Malicious Deep Links

Vulnerability ID: CVE-2026-43944
CVSS Score: 9.4
Published: 2026-05-08

CVE-2026-43944 is a critical vulnerability in the electerm client that allows for arbitrary local code execution. The application insecurely parses deep link arguments and merges untrusted JSON directly into the core session configuration. This enables attackers to override internal state variables, hijacking the application's execution flow to spawn malicious local binaries.

TL;DR

A critical flaw in electerm (< 3.8.15) allows attackers to execute arbitrary local binaries via crafted electerm:// URIs or CLI flags. The application insecurely merges user-provided JSON payloads into the main session configuration, enabling protocol and executable hijacking.

⚠️ Exploit Status: POC

Technical Details

CVSS v4.0: 9.4 (Critical)
EPSS Score: 0.00144 (0.14%)
CWE IDs: CWE-20, CWE-94, CWE-829
Attack Vector: Network (via URI handler)
Exploit Status: Proof of Concept (PoC)
Privileges Required: None
User Interaction: Required

Affected Systems

electerm 3.0.6 - 3.8.14
electerm: >= 3.0.6, < 3.8.15 (Fixed in: 3.8.15)

Code Analysis

Commit: 8a6a179

Implemented OPTS_DENY_LIST to prevent overriding 'type' and 'host' via deep links.

Commit: a79e06f

Added check to reject execution paths containing '..' sequence.

Mitigation Strategies

Upgrade electerm to version 3.8.15 or later.
Unregister the electerm:// protocol handler in the operating system registry or application settings.
Implement Endpoint Detection and Response (EDR) rules to monitor electerm child process creation.
Conduct user training to prevent interaction with untrusted deep links.

Remediation Steps:

Identify all hosts running electerm versions between 3.0.6 and 3.8.14.
Deploy the 3.8.15 update package via centralized endpoint management tools.
Verify the update installation by checking the application version string.
If patching is delayed, execute scripts to remove the electerm:// URI handler association on affected systems.

References

Read the full report for CVE-2026-43944 on our website for more details including interactive diagrams and full exploit analysis.

GHSA-7HGR-XVRR-XPW3: GHSA-7HGR-XVRR-XPW3: Session Persistence After Password Change in Nhost hasura-auth

CVE Reports — Fri, 08 May 2026 18:10:29 +0000

GHSA-7HGR-XVRR-XPW3: Session Persistence After Password Change in Nhost hasura-auth

Vulnerability ID: GHSA-7HGR-XVRR-XPW3
CVSS Score: 7.5
Published: 2026-05-08

A critical session management vulnerability in Nhost's authentication service allows attackers to maintain unauthorized access following a password reset. The password update operation fails to invalidate existing refresh tokens in the database, violating standard session revocation principles and rendering password changes ineffective as an incident response measure.

TL;DR

Nhost's hasura-auth component fails to clear active refresh tokens upon a password change. Attackers holding stolen tokens can continue generating valid access tokens indefinitely.

⚠️ Exploit Status: POC

Technical Details

CWE ID: CWE-613: Insufficient Session Expiration
Attack Vector: Network (Requires stolen refresh token)
Estimated CVSS: 7.5 (High)
Impact: Persistent unauthorized access post-credential rotation
Exploit Status: Conceptually straightforward, requires prerequisite compromise
Patch Status: Fixed in PR #4192

Affected Systems

Nhost hasura-auth backend service
Nhost nhost-js client SDK
hasura-auth: < PR #4192 (Fixed in: PR #4192)
nhost-js: < PR #4192 (Fixed in: PR #4192)

Code Analysis

Commit: 52c7066

Atomic deletion of refresh tokens via CTE upon password update

Mitigation Strategies

Upgrade the Nhost backend services to a release subsequent to PR #4192.
Upgrade the nhost-js SDK to incorporate the updateSessionFromResponseMiddleware.
Implement continuous monitoring on the auth.refresh_tokens table for anomalous long-lived entries.
Educate users to manually terminate active sessions if native global logout is unsupported in their current deployment.

Remediation Steps:

Review the current deployment version of hasura-auth and nhost-js.
Update the backend Go service with the CTE SQL modifications to ensure atomic token deletion.
Update all frontend clients leveraging the nhost-js SDK to force local storage clearing on password changes.
Run a manual cleanup script targeting the auth.refresh_tokens table for any accounts modified prior to the patch application.

References

Read the full report for GHSA-7HGR-XVRR-XPW3 on our website for more details including interactive diagrams and full exploit analysis.

CVE-2026-6321: CVE-2026-6321: Path Traversal in fast-uri via Improper Normalization Order

CVE Reports — Fri, 08 May 2026 17:40:29 +0000

CVE-2026-6321: Path Traversal in fast-uri via Improper Normalization Order

Vulnerability ID: CVE-2026-6321
CVSS Score: 7.5
Published: 2026-05-08

The fast-uri library (versions ≤ 3.1.0) contains a high-severity path traversal vulnerability due to an order-of-operations flaw during URI normalization. The library incorrectly decodes percent-encoded path separators (%2F) and dot segments (%2E) prior to applying dot-segment removal algorithms, allowing attackers to bypass path-based access controls and filters.

TL;DR

fast-uri ≤ 3.1.0 decodes percent-encoded URI characters before running path normalization algorithms. This allows attackers to use payloads like %2e%2e to bypass security filters and perform path traversal attacks. Upgrade to version 3.1.1 to implement context-aware decoding.

⚠️ Exploit Status: POC

Technical Details

CWE ID: CWE-22
Attack Vector: Network
CVSS Score: 7.5 (High)
EPSS Score: 0.00030
Exploit Status: Proof of Concept Available
CISA KEV: Not Listed

Affected Systems

fast-uri <= 3.1.0
Node.js applications utilizing fast-uri for request validation
API Gateways and proxies dependent on fast-uri for routing
fast-uri: <= 3.1.0 (Fixed in: 3.1.1)

Code Analysis

Commit: 876ce79

Fix path traversal by introducing context-aware percent-encoding and path normalization.

Exploit Details

Vulnerability Report Example Payload: Proof of concept payload demonstrating path filter bypass using %2e%2e.

Mitigation Strategies

Upgrade fast-uri to version 3.1.1 or higher.
Deploy WAF rules to block URIs containing percent-encoded dot (%2e) or slash (%2f) sequences.
Ensure path-based authorization checks occur on the fully normalized URI, not the raw input.

Remediation Steps:

Audit application dependencies using npm audit or yarn audit to identify instances of fast-uri <= 3.1.0.
Update the package.json file to require fast-uri version 3.1.1.
Run the package manager update command to fetch the patched library and regenerate the lockfile.
Deploy the updated application and monitor logs for previously blocked traversal attempts.

References

Read the full report for CVE-2026-6321 on our website for more details including interactive diagrams and full exploit analysis.

GHSA-8G7G-HMWM-6RV2: GHSA-8g7g-hmwm-6rv2: Path Traversal, SSRF, and Information Exposure in n8n-mcp

CVE Reports — Fri, 08 May 2026 17:10:29 +0000

GHSA-8g7g-hmwm-6rv2: Path Traversal, SSRF, and Information Exposure in n8n-mcp

Vulnerability ID: GHSA-8G7G-HMWM-6RV2
CVSS Score: 8.5
Published: 2026-05-08

Multiple high-severity vulnerabilities were identified in the n8n-mcp package prior to version 2.50.1. These vulnerabilities include a Path Traversal flaw in the API client, a Server-Side Request Forgery (SSRF) bypass via redirect-following, and an Information Exposure vulnerability in the telemetry service. Collectively, these flaws permit credential theft, internal network access, and the leakage of sensitive workflow configurations.

TL;DR

Versions of n8n-mcp before 2.50.1 suffer from path traversal in API path construction, SSRF via uncontrolled redirect following, and plain-text exposure of sensitive API keys in telemetry data. The vendor patched these issues in version 2.50.1.

⚠️ Exploit Status: POC

Technical Details

Vulnerability IDs: GHSA-8g7g-hmwm-6rv2, AIKIDO-2026-10739
Primary CWEs: CWE-22, CWE-918, CWE-200, CWE-212
Attack Vector: Network
Estimated CVSS: 8.5 (High)
Exploit Status: Proof of Concept available
Patched Version: 2.50.1

Affected Systems

n8n-mcp API Client
n8n-mcp Webhook Triggers
n8n-mcp Telemetry Service
n8n-mcp: < 2.50.1 (Fixed in: 2.50.1)

Code Analysis

Commit: 1cfe9c6

Security patch addressing path traversal, SSRF, and telemetry leakage vulnerabilities in n8n-mcp.

Mitigation Strategies

Upgrade the n8n-mcp package to version 2.50.1 or later.
Implement egress network filtering to block connections to cloud metadata endpoints (169.254.169.254).
Rotate any API keys or credentials that were stored in historical telemetry logs.
Audit custom webhook handlers to ensure HTTP redirect following is explicitly disabled.

Remediation Steps:

Identify all deployments of the n8n-mcp package within the environment.
Execute package manager updates to pull version 2.50.1.
Restart the affected Node.js services to apply the updated dependencies.
Review historical telemetry logs and purge any entries containing unredacted workflow mutation payloads.
Rotate credentials accessed by the application prior to the patch application.

References

Read the full report for GHSA-8G7G-HMWM-6RV2 on our website for more details including interactive diagrams and full exploit analysis.