proj-sync-doc
Pass
Audited by Gen Agent Trust Hub on Sep 23, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill utilizes local shell commands (
git log,git show,git diff) to inspect repository history and code changes. These are standard developer tools used for its primary function. - [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted external data which presents a theoretical attack surface for indirect instructions.
- Ingestion points: Data is ingested from the local repository through git commit messages and code diffs (SKILL.md, Step 1).
- Boundary markers: The skill enforces a mandatory user confirmation step in the
强制规则(Mandatory Rules) and Step 3, requiring the agent to present a change list and wait for approval before proceeding. - Capability inventory: The skill has the capability to write to the local filesystem (updating files in
docs/and unit test directories) and execute local git commands. - Sanitization: Sanitization is primarily handled via human-in-the-loop review of the change points before any modifications occur, reducing the risk of the agent obeying instructions embedded in code or commit history.
Audit Metadata