audit-frontend
Pass
Audited by Gen Agent Trust Hub on Jul 21, 2026
Risk Level: SAFEPROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [COMMAND_EXECUTION]: The skill facilitates the execution of local shell commands as part of its auditing process.
- Evidence: In
references/review-checklist.md, the agent is instructed to "Run only repository-defined non-mutating format checks, lint, typecheck, focused tests, builds, and generated-route checks". These commands are sourced from the target repository being audited. - [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it ingests and prioritizes untrusted documentation and instructions from the analyzed repository.
- Ingestion points: The skill reads
AGENTS.md,CLAUDE.md, and other project documentation to determine repository guidance, which ranks higher in priority than the skill's own rules (SKILL.md,references/architecture-and-ownership.md). - Boundary markers: The instructions do not define delimiters or "ignore embedded instructions" warnings for the content read from the repository.
- Capability inventory: The agent has access to command execution (shell scripts) and browser automation tools (
references/review-checklist.md). - Sanitization: No sanitization or validation of the ingested repository-specific rules is described in the workflow.
Audit Metadata