gf-review
Pass
Audited by Gen Agent Trust Hub on Aug 29, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill uses local shell commands including
git status,git ls-files,find,rg(ripgrep), andopenspecto determine the review scope and project state. These are standard utilities used for their intended purpose within a development environment. - [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted content from the repository, including project source files and the
CLAUDE.mdspecification file, which it treats as the "single source of truth." This creates an attack surface where malicious instructions embedded in the code or documentation could attempt to influence the agent's review logic. - Ingestion points: Reads
CLAUDE.mdand files identified in the repository status (SKILL.md). - Boundary markers: The instructions do not define specific delimiters or guardrails to separate data content from the agent's instructions during the review process.
- Capability inventory: The skill has the ability to execute shell commands and call other skills like
goframe-v2(SKILL.md). - Sanitization: There is no evidence of content sanitization for the files being reviewed.
Audit Metadata