codepatrol-autofix-deploynote
Pass
Audited by Gen Agent Trust Hub on Oct 9, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data from external sources which could contain malicious instructions.
- Ingestion points: Pull Request summaries and human comments are fetched and parsed by the agent as specified in
SKILL.md. - Boundary markers: The skill utilizes a predefined Markdown template (
DEPLOY-NOTE-TEMPLATE.md) and HTML markers (<!-- deploy-notes -->) to structure its output, but it lacks specific instructions for the agent to ignore or escape instructions embedded within the ingested PR content. - Capability inventory: The skill has the ability to read PR data and create or update comments on GitHub repositories.
- Sanitization: No sanitization of external content is performed; the instructions explicitly require the agent to quote commands and definitions verbatim from the PR descriptions, which could allow a malicious actor to inject content into the final deploy note.
Audit Metadata