fli-review

Pass

Audited by Gen Agent Trust Hub on Jul 20, 2026

Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill instructs the agent to execute shell commands to verify the state of the repository:
  • git log --oneline -30: Used to review recent project history.
  • scripts/check-fast.sh: Executes a local repository script to perform health checks.
  • npm --prefix frontend run build: Executes a standard build command to verify frontend claims.
  • [PROMPT_INJECTION]: The skill possesses an indirect prompt injection surface as it processes data from the local repository that could potentially be influenced by external contributors (e.g., via pull requests):
  • Ingestion points: The agent reads markdown documentation (docs/*.md, PRODUCT.md), build logs (docs/references/build-log.jsonl), and queries local SQLite databases (data/fli.db).
  • Boundary markers: There are no explicit delimiters or instructions provided to the agent to disregard potential natural language instructions embedded within these files.
  • Capability inventory: The agent has the capability to execute shell commands (git, npm, and local scripts) based on its analysis of the repo state.
  • Sanitization: The skill does not implement specific sanitization or validation logic for the content of the files it ingests.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 20, 2026, 11:05 AM
Security Audit — agent-trust-hub — fli-review