quantify-impact

Pass

Audited by Gen Agent Trust Hub on Sep 11, 2026

Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
  • [COMMAND_EXECUTION]: The skill executes local scripts such as scripts/resolve-pr-base.sh and requires the agent to run user-defined benchmark or test commands (e.g., bun test) to quantify impact.
  • [EXTERNAL_DOWNLOADS]: The instructions mandate installing dependencies from lockfiles for both the base and candidate code trees, which involves network access and potential execution of untrusted lifecycle hooks during the installation process.
  • [INDIRECT_PROMPT_INJECTION]:
  • Ingestion points: The skill ingests code and configuration from PR branches and the local working tree (REFERENCE.md).
  • Boundary markers: No explicit boundary markers or instructions are provided to prevent the agent from following malicious prompts that might be embedded in the code or PR descriptions being measured.
  • Capability inventory: The skill has the capability to execute shell commands, run benchmarks, and interact with the filesystem (SKILL.md, REFERENCE.md).
  • Sanitization: There is no evidence of sanitization for the commands or scripts loaded from the repository being evaluated.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 11, 2026, 02:00 AM
Security Audit — agent-trust-hub — quantify-impact