changelog
Pass
Audited by Gen Agent Trust Hub on Oct 23, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill provides instructional content for writing blog posts and does not contain any executable code, suspicious network operations, or persistence mechanisms.
- [INDIRECT_PROMPT_INJECTION]: The skill identifies an ingestion surface where the agent is instructed to read external PR descriptions, issues, and commit messages to inform blog content. Ingestion points: The agent gathers context from external GitHub PRs and issues (Rule 1a). Boundary markers: The instructions do not specify any delimiters or safety warnings for this external content. Capability inventory: The skill is limited to prose generation for static blog files and does not call any dangerous tools like subprocesses or network writers. Sanitization: There are no instructions for sanitizing or validating the external data. This is assessed as safe because the impact is limited to the generated text of the blog post.
Audit Metadata