@n8n-as-code/skills
The AI review rates the findings as likely false positive, but the risk score (68/100) still counts them. The review covers version 1.1.5, not the analysed version 2.4.1.
Analysis record
- Analysed
- 2 months ago 11 September 2026
- Version
- v2.4.1
- Artifact
- SHA256 2864d5cb…efd77e
- Detail
- Full list
Evidence ledger
Ranked by severity · findings with a source location link to the code viewer
Finding Categories
AI Security Report
AI Security Review
Evidence context: threat category none; evidence quality moderate.
Tool-Poisoning Assessment
The CVEQ scanner found zero tool-poisoning findings for @n8n-as-code/skills. This is the most significant indicator: there is no evidence of hidden AI manipulation directives, no XML-style instruction tags, no Unicode steganography, and no tool descriptions containing covert instructions aimed at AI agents. The package defines legitimate MCP tool skills for n8n workflow automation, as stated in its description.
Credential and Network Access
There are zero credential-access findings. The package does not read sensitive paths like .ssh/, .aws/credentials, .kube/config, or target specific secret names (GITHUB_TOKEN, AWS_SECRET_ACCESS_KEY, etc.). This is normal for an MCP server that doesn't need to harvest credentials.
Two network findings exist: NET-FETCH-package/dist/commands/skills-commander.js-552 and NET-FETCH-package/dist/commands/workflows.js-95. Both are in the dist/ folder (bundled production code) and show generic fetch detection without suspicious domain information. The findings do not indicate exfiltration architecture because: (1) no credential-access findings coexist with these network calls, and (2) no suspicious external domains are flagged. These fetch calls are consistent with legitimate MCP server behavior for API communication.
Dependencies
All six dependency findings are standard npm packages: commander@^11.1.0, flexsearch@^0.8.212, zod@^3.22.4, @modelcontextprotocol/sdk@^1.27.1, @n8n-as-code/[email protected], and chalk@^4.1.2. These are legitimate development dependencies for building an MCP server with CLI commands and schema validation.
Strongest Counterargument
The package shows 0 users and lists "GitHub Actions" as the developer name, which could indicate an automated or unvetted publishing workflow. However, the absence of any malicious patterns—no tool poisoning, no credential theft, no suspicious domains—means this counterargument does not change the conclusion. New packages with automated publishing are common in the MCP ecosystem.
Conclusion
This package exhibits no evidence of malicious behavior. The findings are entirely consistent with a legitimate MCP server library for n8n workflow automation.
Key Reasons
- Zero tool-poisoning findings detected
- Zero credential-access findings
- Network findings are generic fetch calls in bundled dist/ code with no suspicious domains
- All dependencies are legitimate npm packages for MCP server development
- No malware signatures or obfuscation detected
False Positive Considerations
- Generic fetch detection in bundled dist/ code
- Dependency findings from standard npm packages
- Scanner noise from MCP server architecture patterns
Reviewed 2026-04-27 · version 1.1.5; recommended action: no action; model confidence 85%.
n8n version history
Risk trend by version
6 analyzed versions. Each point is the latest successful scan for that version; failed zero-score scans are hidden. Dates: First seen by Risky Plugins.
Pick any point on the chart, then open the Source tab to read that version's code.
Source Code Not Available
Source code is not available for this version of the extension.