mece-plan-review

Pass

Audited by Gen Agent Trust Hub on May 19, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
  • [SAFE]: Comprehensive analysis of the 8 skill files found no evidence of malicious intent, credential theft, or hidden behaviors. The skill logic is focused on software engineering quality assurance.
  • [COMMAND_EXECUTION]: The skill uses git and the GitHub CLI (gh) to resolve the repository organization and fetch a list of related repositories. These commands are executed locally within the user's development environment to provide context for the review analysts.
  • [EXTERNAL_DOWNLOADS]: The Black Box analyst (bb-analyst.md) utilizes WebFetch to access technical specifications and documentation from public sources like RFCs or official vendor docs, which is appropriate for its role in specification verification.
  • [PROMPT_INJECTION]: The skill ingests untrusted content from plan files and analysis documents. The main agent interpolates this data into subagent prompts using descriptive boundary markers (e.g., 'プランファイル:'). The system leverages subagents with file access and search capabilities to perform its review, though it does not implement formal sanitization of the input text.
Audit Metadata
Risk Level
SAFE
Analyzed
May 19, 2026, 08:36 AM
Security Audit — agent-trust-hub — mece-plan-review