platform-docs-writer
Pass
Audited by Gen Agent Trust Hub on Oct 1, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data from external repository files (such as
Dockerfile,compose.yaml, and.env.example) to generate its output. While the instructions are highly structured and focused on operational data, there is a potential for instructions hidden within these files (e.g., in comments or configuration descriptions) to be transcribed into the final documentation if not properly sanitized. - Ingestion points: Files within the analyzed repository (read via the agent's file system tools).
- Boundary markers: The skill uses strict templates for output, but does not provide specific instructions to the agent on how to escape or sanitize content extracted from external files.
- Capability inventory: The skill can create and modify files (
README.md,docs/*.md) and potentially perform network diagnostics if suggested by the agent. - Sanitization: None explicitly mentioned beyond specific rules for masking credentials.
- [SAFE]: The skill incorporates strong security best practices regarding credential management. It explicitly forbids the agent from copying actual secrets or private keys into documentation, even if they are found in the repository, requiring instead that the agent use placeholders or describe the required format.
Audit Metadata