New: Skill bounties are live. Post a request, fund the bounty, and creators compete for 7 days to build it -> See open bounties
THE AGENSI STORE
3 skills found
An adversarial gate that audits an AI eval or test suite — LLM-judge rubrics, datasets, regression tests, metrics — for gameable criteria, data leakage, missing edge cases, and non-determinism, then returns one PASS/REVISE/FAIL verdict.
Design and write the eval suite for your LLM-powered feature — the metrics that match your failure modes, a golden dataset plan with starter cases, anchored rubrics, LLM-as-judge prompts with the known bias mitigations, and pass/fail gates wired for CI.
by Shandra
Audits image and video datasets for exact and near duplicates, cross-split leakage, class imbalance, suspected labeling errors, annotation inconsistencies, blur, low resolution, corruption, compression artifacts, visual shortcuts, source bias, insufficient diversity, synthetic-data weaknesses, and deployment-domain gaps. Produces an evidence-based quality report with class-balance analysis, labeling risks, severity-ranked findings, and a prioritized remediation and re-audit plan.