review-acid
Pass
Audited by Gen Agent Trust Hub on Oct 1, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill uses local Git commands to identify the scope of files to review.
- Evidence: Executes
git diff --name-only --diff-filter=d HEADandgit symbolic-ref -q --short refs/remotes/origin/HEADto determine which files have changed. - [INDIRECT_PROMPT_INJECTION]: The skill analyzes source code from the repository, which is a surface for indirect prompt injection if the code contains malicious instructions.
- Ingestion points: Ingests file content identified via Git diffs or user-provided file arguments.
- Boundary markers: The skill does not implement specific boundary markers or delimiters when interpolating file content into its analysis prompt, though it instructs the agent to be 'evidence-gated'.
- Capability inventory: The skill has read access to the local filesystem and Git history; it lacks network access and write permissions in its primary instructions.
- Sanitization: No explicit sanitization or filtering of the ingested source code content is performed before processing.
Audit Metadata