night-market-diagnostics-toolkit

Pass

Audited by Gen Agent Trust Hub on Jul 31, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The check_pinned_versions.py script queries GitHub to verify that tool pins match the latest upstream releases. This is a documented administrative function for maintaining tool currency.
  • [COMMAND_EXECUTION]: The toolkit executes various local Python and Bash diagnostic scripts within the repository environment. These operations are intended for development-time health checks and repository auditing.
  • [PROMPT_INJECTION]: The skill exhibits a surface for indirect prompt injection.
  • Ingestion points: Diagnostic tools read content from SKILL.md files (via validate_budget.py), Markdown files (via check-markdown-links.py), and Python source code files (via check_docstring_quality.py) within the repository.
  • Boundary markers: There are no explicit markers or instructions within the scripts to ignore potential agent instructions embedded in the analyzed data.
  • Capability inventory: The skill possesses the capability to execute shell commands and scripts based on the results of its analysis of the repo data.
  • Sanitization: The provided logic does not demonstrate specific sanitization or escaping of content retrieved from the repository files before processing.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 31, 2026, 01:38 AM
Security Audit — agent-trust-hub — night-market-diagnostics-toolkit