chinese-doc-style
Pass
Audited by Gen Agent Trust Hub on Oct 27, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it processes untrusted external data (such as existing project documentation and user-provided text) and possesses file-writing capabilities to update these documents.\n
- Ingestion points: The skill reads file content from the local file system (e.g.,
docs/architecture.md,README.md) and takes input directly from user messages to perform formatting and rewriting tasks.\n - Boundary markers: There are no explicit boundary markers or delimiters defined in the instructions to isolate the untrusted document content from the agent's internal instruction set.\n
- Capability inventory: The skill is authorized to perform file-write operations, specifically writing updated content back to the source files.\n
- Sanitization: The skill lacks explicit instructions for sanitizing, escaping, or filtering the input data to prevent malicious instructions embedded in the documents from being executed by the agent.
Audit Metadata