Security Audit: 2 finding(s) in mcp #2027
Description
AgentAudit Security Audit Report
| Metric | Value |
|---|---|
| Package | `mcp` |
| Risk Score | 20/100 |
| Result | safe |
| Findings | 2 total (0 critical, 1 high, 1 medium, 0 low) |
Findings Summary
-
HIGH - shell=True with subprocess on Windows (mcp/cli/cli.py:48)
- The CLI module uses subprocess.run with shell=True on Windows when checking for npx command availability and when launching the MCP Inspector. Using shell=True expands attack surface.
-
MEDIUM - Environment variable function filtering may be incomplete (mcp/client/stdio/init.py:63)
- Code filters environment variables starting with '()' to prevent shell function injection, but this only covers bash shellshock-style attacks and may not cover all edge cases.
Full Report
View the complete audit report with details, evidence, and remediation guidance:
AgentAudit Report
This audit was performed automatically by AgentAudit, the security registry for AI agent packages. If you believe any finding is incorrect, you can dispute it on the platform.