MCP Registry

@kernlang/mcp-server

by cukasn
96d744d5-97bd-56a4-bf3d-9b2503cffd96 | Store v4.5.0
17/ 100
MINIMAL risk
No change since v4.1.0
AI review verdict
No high-risk signal observed

Based on the RiskyPlugins AI security review of the observed evidence. The review covers version 3.3.9, not the analysed version 4.5.0.

Analysis record

Analysed
4 months ago 8 July 2026
Version
v4.5.0
Artifact
SHA256 62251514…4837aa
Detail
Full list

No code findings

All 6 counted findings are network indicators. They are listed in the Network tab.

Finding Categories

6
IoC Indicators

MCP Server Analysis

MCP servers expose tools and resources to AI assistants. Unlike browser extensions, they run as standalone processes with direct system access. Tool definitions are analyzed for prompt injection, data exfiltration, and tool poisoning patterns.

AI Security Report

AI Security Review

Evidence context: threat category none; evidence quality strong.

Tool-Poisoning Assessment

Zero tool-poisoning findings were detected. The findings_summary shows "tool-poisoning":"0", meaning no hidden AI manipulation directives, no XML-style instruction tags, and no invisible Unicode characters encoding instructions. This is a strong indicator of benign behavior.

Credential and Network Access

No credential harvesting or suspicious network activity was found. The findings_summary shows "secret":"0" and "network":"0", indicating the package does not read sensitive paths like .ssh/, .aws/credentials, or .kube/config, and does not make HTTP calls to unknown domains. The only URLs detected are legitimate project infrastructure:

  • XIOC-URL-https://github.com/KERNlang/kern - Official KERN language repository
  • XIOC-URL-https://kernlang.dev - Official project website
  • XIOC-URL-https://www.gnu.org/licenses/ - GNU license page
  • XIOC-DOMAIN-fsf.org - Free Software Foundation
  • XIOC-DOMAIN-github.com - GitHub

IOC Findings Are False Positives

All 30 IOC findings are clearly false positives from the XIOC extractor's known garbage patterns:

  • Source map files misidentified as domains: XIOC-DOMAIN-drifts.map, XIOC-DOMAIN-srv.tools.map, XIOC-DOMAIN-result.configsmissing.map, XIOC-DOMAIN-index.js.map
  • Property access chains misread as domains: XIOC-DOMAIN-server.name, XIOC-DOMAIN-server.trust, XIOC-DOMAIN-srv.name

These match the documented false-positive patterns: "Property access chains misread as domains: b.call, h.next, g.id" and source map files ending in .map.

Code-Smell Findings

The 12 code-smell findings (severity: low) are noise per the guidelines. These are YARA rules matching basic Node.js patterns and do not indicate malicious behavior.

Strongest Counterargument

The 30 IOC findings and 45 total findings could suggest suspicious activity. However, the guidelines explicitly state "IoC COUNT alone is meaningless" and "Do NOT anchor on risk_score. Derive your verdict from the actual findings." Every IOC finding is a documented false-positive pattern (source maps, property access), and there are zero findings in the categories that matter: tool-poisoning, secret access, and suspicious network activity. The package serves a legitimate purpose (KERN language compilation) with URLs pointing to its official infrastructure.

Conclusion

This MCP server shows no evidence of malicious behavior. All findings are explained false positives from the scanner's known noise sources.

Key Reasons

  • Zero tool-poisoning findings
  • Zero credential/secret access findings
  • Zero suspicious network activity findings
  • All IOC findings are documented false-positive patterns
  • URLs point to legitimate KERN language project infrastructure

False Positive Considerations

  • Source map files misidentified as domains (.map extensions)
  • Property access chains misread as domains (server.name, srv.name)
  • Code-smell findings from basic Node.js patterns
  • Legitimate project URLs (GitHub, kernlang.dev, fsf.org)

Reviewed 2026-04-27 · version 3.3.9; recommended action: suppress false positive; model confidence 85%.

MCP version history

Risk trend by version

20 analyzed versions. Each point is the latest successful scan for that version; failed zero-score scans are hidden. Dates: First seen by Risky Plugins.

Selected
17
Change since first
-13
Change from previous
No change
Versions:
First analyzed version
3.3.6
First seen Apr 21, 2026
Risk range
17 to 30
Across analyzed versions
Latest analyzed version
4.5.0
First seen Jul 8, 2026
Selected version
minimal
Version
v4.5.0
First seen 4 months ago
Risk score
17
Findings
6
Change vs previous
No change

Pick any point on the chart, then open the Source tab to read that version's code.

Source Code Viewer

Source Code Not Available

Source code is not available for this version of the extension.

About This Extension

KERN MCP server — compile, review, and analyze .kern files via MCP

Frequently Asked Questions