mega-code-status
Pass
Audited by Gen Agent Trust Hub on Apr 10, 2026
Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: Executes shell commands via the Bash tool to verify authentication and retrieve pipeline status using the 'uv' package manager and the 'mega-code' CLI tool.
- [PROMPT_INJECTION]: Ingests untrusted metadata (descriptions and headers) from pending files located in the tool's data directories. Evidence: Ingestion points include 'SKILL.md' and strategy files in '~/.local/share/mega-code/data/'; capability inventory includes subprocess execution via Bash; no boundary markers are used to separate ingested content from instructions; sanitization is limited to character truncation via 'head'.
Audit Metadata