webflux-test-reviewer
Pass
Audited by Gen Agent Trust Hub on Mar 16, 2026
Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill executes 'git log --oneline' and 'git branch -a' within candidate directories to evaluate commit history and branching strategies. These commands are localized to the assessment project and are functional requirements for the evaluation process.
- [PROMPT_INJECTION]: The skill presents an indirect prompt injection surface as it ingests and processes untrusted data from candidate repositories. 1. Ingestion points: Candidate source code, configuration files, and README documentation located in root subdirectories (SKILL.md). 2. Boundary markers: None; the skill lacks specific instructions to the agent to disregard instructions or overrides embedded within the files under review. 3. Capability inventory: The agent can read filesystem content, execute git commands, and write evaluation reports (EVALUATION.md). 4. Sanitization: There is no explicit sanitization or filtering of the candidate-provided content before it is analyzed by the agent.
Audit Metadata