agile-refinement
Pass
Audited by Gen Agent Trust Hub on May 12, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill's primary function is to analyze local text-based artifacts (Markdown files and code diffs) to provide qualitative feedback and linting results.
- [DATA_EXPOSURE]: The instructions explicitly direct the agent to check for and report the presence of raw secrets, provider keys, tokens, or credentials within planning artifacts, which serves as a security improvement for the user's workflow.
- [INDIRECT_PROMPT_INJECTION]: The skill possesses an ingestion surface for indirect prompt injection as it reads external data (planning files and code diffs). However, the risk is negligible as the skill's capabilities are limited to generating inline text reports, with no evidence of dangerous tool use, network exfiltration, or command execution.
- [COMMAND_EXECUTION]: While the skill accepts input via
$ARGUMENTS, it uses this data only to determine the operational mode (planning vs. code review) and does not interpolate it into shell commands or executable contexts.
Audit Metadata