bansos-develop-feature

Pass

Audited by Gen Agent Trust Hub on Jul 29, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [PROMPT_INJECTION]: The skill processes project data, including Markdown and JSON files provided by contributors (src/lib/data/bansos/).
  • Ingestion Points: File reads of listing and contributor data (e.g., manifest.json, README.md).
  • Capability Inventory: Includes command execution via npm scripts (npm run build, check, lint) and filesystem write access to the repository.
  • Boundary Markers: No explicit delimiters or instructions are provided to the agent to disregard instructions potentially embedded within these data files.
  • Sanitization: No specific data validation or sanitization steps for the content of these files are mentioned beyond standard build checks.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 29, 2026, 01:40 PM
Security Audit — agent-trust-hub — bansos-develop-feature