backend-to-frontend-handoff-docs
Pass
Audited by Gen Agent Trust Hub on Feb 17, 2026
Risk Level: SAFEPROMPT_INJECTIONNO_CODE
Full Analysis
- Indirect Prompt Injection (LOW): The skill possesses a vulnerability surface for indirect prompt injection as it ingests and processes untrusted data (source code).
- Ingestion points: The skill is designed to analyze completed backend code, including endpoints, controllers, services, DTOs, and validation logic, as specified in
SKILL.md. - Boundary markers: There are no defined boundary markers or instructions telling the agent to treat strings or comments within the analyzed code as data only and to ignore any embedded instructions.
- Capability inventory: The skill has the capability to write files to the local file system, specifically to the
.claude/docs/ai/directory. - Sanitization: There is no mention of sanitizing or escaping the content extracted from the source code before it is formatted into the final markdown document.
- NO_CODE (SAFE): The skill consists entirely of markdown instructions and documentation. No executable scripts (Python, Node.js, etc.) are included in the skill package.
Audit Metadata