modelshow

Blind multi-model comparison with architecturally guaranteed de-anonymization. Trigger with "mdls" or "modelshow" for double-blind evaluation of AI model res...

The skill appears to implement the described multi-model blind-judging workflow, but take these precautions before installing or running it: 1) Understand the anonymity tradeoff: despite wording claiming the orchestrator never sees placeholders, the anonymize step returns an anonymization_map (placeholder→model) to the orchestrator — keep that map secret or modify the workflow if you require stronger guarantees. 2) Review and edit config.json: remove any model aliases you don't want queried and set outputDir to a safe location. 3) Be aware the skill will read your OpenClaw config (~/.openclaw/openclaw.json) to resolve model aliases and will write reports to your home directory; if that file contains sensitive info, inspect save_results.py and blind_judge scripts first. 4) External-content behavior: the SKILL.md says the agent will fetch referenced URLs or files and prepend them to prompts — if you want to avoid exposing local or remote data, disable this behavior or ensure you only run the skill on non-sensitive prompts. 5) If you require the stronger property that nobody except the judge can deanonymize, either remove the return of anonymization_map from the anonymize phase or ensure anonymization and finalize are executed inside a trusted atomic environment that never exposes the map to the orchestrator. 6) If you are unsure, run the skill on harmless test prompts first and manually inspect the outputs and saved files. If you want, I can point out the precise lines to change to avoid returning the anonymization map to the orchestrator or to prevent reading ~/.openclaw/openclaw.json.

Type: OpenClaw Skill Name: modelshow Version: 1.0.1 ModelShow is a legitimate tool for double-blind AI model evaluation and benchmarking. The skill bundle contains Python scripts (`judge_pipeline.py`, `save_results.py`, and `blind_judge_manager.py`) that handle the anonymization of model responses and the local persistence of results in Markdown and JSON formats. The code uses standard libraries, implements cryptographically secure shuffling for unbiased judging, and restricts file operations to the user's local OpenClaw workspace. No evidence of data exfiltration, malicious command execution, or prompt-injection attacks was found.