codebase-scanning

Pass

Audited by Gen Agent Trust Hub on Apr 14, 2026

Risk Level: SAFEPROMPT_INJECTIONNO_CODE
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and analyze content from untrusted codebase files, creating a vulnerability surface for indirect prompt injection where malicious instructions could be embedded in manifest files or source code.\n
  • Ingestion points: The agent is instructed to read various project files across the entire repository, including package manifests (e.g., package.json, requirements.txt, build.gradle) and source code (routes, components, views) to extract context.\n
  • Boundary markers: The instructions do not define specific delimiters or include warnings for the agent to ignore instructions found within the scanned data.\n
  • Capability inventory: The skill utilizes file system read capabilities to perform scans and is instructed to use a 'Write' tool to save validated findings back to the codebase.\n
  • Sanitization: There are no requirements for sanitizing or validating the content extracted from the scanned files before processing by the LLM.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 14, 2026, 12:18 PM
Security Audit — agent-trust-hub — codebase-scanning