review-swarm

Pass

Audited by Gen Agent Trust Hub on Jun 24, 2026

Risk Level: SAFECOMMAND_EXECUTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill uses git diff and git diff --cached to retrieve codebase changes for review. These are read-only operations restricted to the local repository scope.
  • [PROMPT_INJECTION]: The skill processes untrusted external data (git diffs and project documentation) which presents an indirect prompt injection surface.
  • Ingestion points: Git diff outputs and local files like AGENTS.md identified in SKILL.md Step 1.
  • Boundary markers: No explicit boundary instructions or delimiters are provided to the sub-agents to separate intent packets from the untrusted content.
  • Capability inventory: Sub-agents are granted permission to run git diff and read files but are explicitly prohibited from using tools like apply_patch or performing any workspace-mutating actions.
  • Sanitization: No data sanitization or input filtering is performed on the ingested content before it is passed to sub-agents.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 24, 2026, 11:18 PM
Security Audit — agent-trust-hub — review-swarm