feature-tech-design
Pass
Audited by Gen Agent Trust Hub on Sep 29, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes external files such as
docs/{功能名称}/1_需求文档.mdanddocs/{功能名称}/prototypes/*.html. These files, which may be authored by third parties, represent a surface for indirect prompt injection. If these files contain malicious instructions, the agent might incorporate them into its design decisions or output. - Ingestion points: The agent reads markdown requirement files and HTML prototypes from the local filesystem (
SKILL.md). - Boundary markers: No explicit boundary markers or "ignore embedded instructions" warnings are used when interpolating file content into the prompt context.
- Capability inventory: The skill has the capability to write files (
docs/{功能名称}/2_技术方案.md) and read various project files (SKILL.md). - Sanitization: There is no evidence of sanitization or filtering applied to the ingested external content before it is processed.
Audit Metadata