frontend-review-triage

Pass

Audited by Gen Agent Trust Hub on Jul 9, 2026

Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill instructs the agent to execute a local shell script scripts/audit-triage.sh with a user-provided repository path as an argument. This is the core mechanism for generating the triage report.\n- [PROMPT_INJECTION]: The skill processes untrusted data from repository files (README.md, package.json) and GitHub issue summaries. While this creates a surface for indirect prompt injection, the impact is limited by the skill's defined boundaries and its local-only reporting focus.\n
  • Ingestion points: package.json, README.md, .github/workflows/, and GitHub issue metadata via gh issue list.\n
  • Boundary markers: Absent; the agent is not explicitly instructed to ignore instructions embedded within the files it reads.\n
  • Capability inventory: File system write access (limited to .frontend-review/ directory) and local shell script execution.\n
  • Sanitization: No explicit sanitization or validation steps are defined for the ingested repository content.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 9, 2026, 03:37 AM
Security Audit — agent-trust-hub — frontend-review-triage