Skills
Navigate
skills/martinacostadev/claude-react-agents/review/Gen Agent Trust Hub

review

Pass

Audited by Gen Agent Trust Hub on Sep 27, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data from external sources including GitHub Pull Request bodies/files via gh pr view and local code changes via git diff as described in SKILL.md. There are no boundary markers or instructions to the agent to ignore embedded commands within the ingested content. Combined with the Bash tool capability, a malicious PR could attempt to influence the agent's behavior.
  • [COMMAND_EXECUTION]: The skill interpolates user-supplied input directly into shell commands (gh pr view $ARGUMENTS) in SKILL.md. This presents a surface for command injection if the environment does not provide automatic argument sanitization.
  • [INDIRECT_PROMPT_INJECTION_EVIDENCE]:
  • Ingestion points: gh pr view output and git diff output (referenced in SKILL.md).
  • Boundary markers: Absent.
  • Capability inventory: Bash, Read, Glob, Grep (allowed-tools in SKILL.md).
  • Sanitization: Absent.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 27, 2026, 10:11 AM