ux-spec
Pass
Audited by Gen Agent Trust Hub on Jun 18, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [SAFE]: No malicious behavior was detected. The skill is designed for documentation generation and uses only safe file system tools for reading and writing markdown files. No network activity or script execution is performed.
- [PROMPT_INJECTION]: The skill possesses an attack surface for indirect prompt injection because it ingests data from local files (docs/PRD.md and docs/ARCHITECTURE.md). While a vulnerability surface exists, the risk is assessed as safe because the agent lacks the capabilities (like network access or shell execution) required to weaponize an injection.
- Ingestion points: Reads from docs/PRD.md, docs/ARCHITECTURE.md, and product-plan-*.md.
- Boundary markers: Absent.
- Capability inventory: Read, Write, Glob, Grep, AskUserQuestion.
- Sanitization: None performed.
Audit Metadata