explain-changes
Pass
Audited by Gen Agent Trust Hub on Oct 22, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill establishes an attack surface by instructing the agent to ingest and analyze potentially untrusted data from multiple sources.
- Ingestion points:
SKILL.mdspecifies that the agent should retrieve and read PR descriptions, linked issues or specs, Git diffs, untracked files, and relevant documentation. - Boundary markers: The instructions lack explicit boundary markers or directives to treat external content as data rather than instructions, which could lead the agent to follow commands embedded in a PR description or code comment.
- Capability inventory: The skill assumes the agent can read local files, access repository-host tools (API/CLI), and optionally run tests or checks to resolve questions.
- Sanitization: There are no requirements for sanitizing, escaping, or filtering external content before the agent processes and explains it.
Audit Metadata