yy-skill-to-prompt
Pass
Audited by Gen Agent Trust Hub on Jul 16, 2026
Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill performs file system write operations. In 'Step 5. 刘建 prompts 文件', the agent is instructed to create or update a file at the path '/prompts/skill-prompts.md' within the repository. This operation is used to persist the generated system prompts for future use.
- [PROMPT_INJECTION]: The skill exhibits an indirect prompt injection attack surface (Category 8). It is designed to ingest and process untrusted external data and transform it into a system prompt. Ingestion points: The skill processes raw content from user-provided skill files (SKILL.md, YAML, etc.). Boundary markers: No explicit instructions are provided for the agent to use delimiters or sanitization techniques to prevent instructions embedded in the input from influencing the output structure. Capability inventory: The skill has the capability to write to the file system (Step 5). Sanitization: No validation or filtering of input content is specified, creating a risk where malicious instructions in the source file could be formatted into a persistent system prompt.
Audit Metadata