setup-agent-loops

Pass

Audited by Gen Agent Trust Hub on Jul 12, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [SAFE]: The skill operates on a human-in-the-loop principle, ensuring that no files are modified and no configurations are set without the user reviewing a draft and providing explicit confirmation.\n- [SAFE]: Instructions explicitly define safety constraints, marking sensitive operations like pushing code or modifying environment variables as actions that require human approval.\n- [PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface as it ingests project files to determine the repository's configuration.\n
  • Ingestion points: Scans package.json, pyproject.toml, Cargo.toml, go.mod, Makefile, and existing agent documentation in SKILL.md Step 1.\n
  • Boundary markers: No explicit delimiters are used to separate ingested file content from agent instructions.\n
  • Capability inventory: Capable of writing to docs/agents/loops.md and modifying CLAUDE.md or AGENTS.md in SKILL.md Step 4.\n
  • Sanitization: The skill relies on the user to review the generated draft before it is written to the filesystem.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 12, 2026, 11:17 AM
Security Audit — agent-trust-hub — setup-agent-loops