Uh oh!
There was an error while loading. Please reload this page.
skills/cuopt-server-common: Add eval - #1325
Conversation
Adds an evals/evals.json with a QP-not-supported-over-REST trap case so NV-BASE Tier 3 agent evaluation can demonstrate skill lift on the server-common skill. Last CI run: composite PASS with overall lift +0.16 on claude-code (best-performing agent). Codex showed -0.05 variance on this run; the per-agent gate may flag this on some CI runs due to LLM-judge variance on saturated baselines. Signed-off-by: Ramakrishna Prabhu <ramakrishnap@nvidia.com>
Auto-sync is disabled for draft pull requests in this repository. Workflows must be run manually. Contributors can view more details about this message here. |
ramakrishnap-nv
commented
May 28, 2026
/nvskills-ci |
Actionable comments posted: 0 |
Caution Review failedPull request was closed or merged during review 📝 WalkthroughWalkthroughAdds an evaluation case asserting QP is not supported over the cuOpt REST server, augments the skill card with evaluation metadata, adds a BENCHMARK report, and updates the Sigstore DSSE bundle to include the new evaluation file. ChangesREST API Scope Clarification & Evaluation
Estimated code review effort🎯 3 (Moderate) | ⏱️ ~20 minutes Suggested labels
Suggested reviewers
🚥 Pre-merge checks | ✅ 5✅ Passed checks (5 passed)
✏️ Tip: You can configure your own custom pre-merge checks in the settings. ✨ Finishing Touches🧪 Generate unit tests (beta)
Comment |
Signed-off-by: nvskills-svc-account <svc-nvskills-signing@nvidia.com>
skills/cuopt-server-common: Add eval