Skill Eval

Skill evaluation framework. Use when: testing trigger rate, quality compare (with/without skill), or model comparison. Runs via sessions_spawn + sessions_his...

This skill is internally consistent with its stated purpose, but take these precautions before running it: - Review SKILL.md and the bundled scripts (especially anything that writes files) so you understand what will be stored under eval-workspace/. - Don't run evaluations against skills or prompts that will surface sensitive credentials or personal data; persisted histories include tool calls and tool results and may capture secrets. - Because the workflow uses sandbox="inherit" and cleanup="keep", spawned subagents inherit the main agent environment and histories are retained — consider running in a disposable/test account or environment if you have any sensitive registrations or tokens available to your agent. - If you need to test locally, create a clean ~/.openclaw/openclaw.json or ensure skills.load.extraDirs points only to safe directories; the resolver reads that file to find skill paths. - The skill does not auto-install anything (fake-tool requires manual copy + gateway restart), and there are no remote downloads in SKILL.md — still, inspect any scripts before executing them in your environment. If you want to proceed safely: run evaluations on a non-production agent, delete eval-workspace/ after reviewing results, and avoid exposing real credentials during tests.

Type: OpenClaw Skill Name: openclaw-skill-eval Version: 1.1.1 The bundle is a comprehensive evaluation framework that utilizes high-risk capabilities, including spawning autonomous sub-sessions via 'sessions_spawn' and executing local Python scripts through 'subprocess' and 'exec' (e.g., in scripts/legacy/run_orchestrator.py). While these behaviors are aligned with the stated purpose of benchmarking AI skills, the framework lacks input sanitization, creating a risk of Remote Code Execution (RCE) if the agent processes untrusted evaluation metadata. Additionally, viewer/generate_review.py starts a local HTTP server and uses 'os.kill' to manage ports, which is aggressive behavior for a skill bundle. No evidence of intentional data exfiltration or backdoors was found, but the broad permissions required for operation warrant a suspicious classification.