linzhe001/evaluate
.agents/skills/evaluate/SKILL.md
Codex wrapper for experiment analysis and decision-making. Use when the user wants metrics interpreted, a stage or iteration report written, and a NEXT_ROUND, DEBUG, CONTINUE, PIVOT, or ABORT recommendation.
$ install --global
skillsauthnpx skillsauth add linzhe001/Harness-Research evaluateInstall this skill globally with one command. Works with Claude Code, Cursor, and Windsurf.
Security Scan Results
3 of 9 scanners reported clean
Some scanners were skipped, did not run, or reported a non-clean status. Review each row below.
3
Scanners Passed
9
Scanners in report
Clean
TrivyContainer and dependency vulnerability scanner
95%
Clean
SemgrepStatic code analysis for vulnerabilities
95%
Clean
mcp-scan (Snyk)Model Context Protocol security validation
95%
Skipped
Snyk (dep)Open source security scanning
50%
Skipped
Socket.devSupply chain security analysis
50%
Skipped
VirusTotalMulti-engine malware detection
50%
Skipped
CrowdStrikeAdvanced threat intelligence
50%
Skipped
OSV-ScannerOpen Source Vulnerability database check
50%
Skipped
OWASP Dep-Check
50%
Last scanned: Apr 17, 2026, 1:58 AM29.9s3 files scanned
SKILL.md
- name:
- evaluate
- description:
- Codex wrapper for experiment analysis and decision-making. Use when the user wants metrics interpreted, a stage or iteration report written, and a NEXT_ROUND, DEBUG, CONTINUE, PIVOT, or ABORT recommendation.
Evaluate
References
Read these first:
../../../.agents/references/workflow-guide.md../../../.agents/references/language-policy.md./references/stage-report.md../../../iteration_log.json../../../PROJECT_STATE.json
When To Use
Use this skill when the user wants training or evaluation results interpreted and turned into a decision.
Required Work
- Parse the relevant logs, metrics, or checkpoint metadata.
- Analyze:
- training stability
- convergence
- overfitting
- gradient and numerical issues
- Resolve the tracked metric set from active iteration context or project state, and compare against baseline metrics and prior iterations using that protocol.
- Produce the canonical report using
./references/stage-report.md. - Recommend exactly one of:
NEXT_ROUND— ordinary improvement round, stay in WF8DEBUG— fixable technical issue, stay in WF8CONTINUE— handoff to orchestrator/WF9, not continue iteratingPIVOTABORT
- If invoked from
$iterate, do not take over stage-transition ownership.
Codex Adaptation
- Treat natural-language requests as the canonical
$evaluateflow or the evaluation sub-step of$iterate. - Use
.agents/state/current_iteration.jsonas the active iteration context path. - Preserve the original decision vocabulary (NEXT_ROUND / DEBUG / CONTINUE / PIVOT / ABORT) and per-iteration reporting behavior, but do not assume a fixed metric family.
- Use
../../../.agents/references/language-policy.mdfor reply language and for localizing natural-language report sections; keep protocol keys and decision tokens in English.
Execution Rule
Follow the local evaluation prompt, stage-report template, and language policy rather than collapsing this into a brief metrics summary.
Related Skills
linzhe001/validate-run
development
VerifiedTrustedCommunity
WF7.5 training pipeline validation. Before entering WF8 iteration, first use Codex to review code for baseline equivalence, then run a 100-step smoke test to verify end-to-end pipeline functionality.
1SKILL.mdUpdated Apr 17, 2026
linzhe001/survey-idea
business
VerifiedTrustedCommunity
WF1 Inspiration survey and gap analysis. Takes the user's research idea, performs literature search, gap analysis, competitor analysis, and feasibility scoring, then outputs Feasibility_Report.md. Use when the user has a new CV research idea that needs a feasibility assessment.
1SKILL.mdUpdated Apr 17, 2026
linzhe001/release
tools
VerifiedTrustedCommunity
WF10 Submission/Release Tool. Multi-scene training, result packaging, filename validation, dry-run submission checks. Used after ablation experiments are complete and before competition submission.
1SKILL.mdUpdated Apr 17, 2026
linzhe001/refine-arch
development
VerifiedTrustedCommunity
WF2 Architecture refinement and MVP design. Reads the feasibility report, analyzes the base codebase architecture, designs plug-and-play new modules, defines the MVP, provides A/B/C alternative plans, and outputs Technical_Spec.md. Use when a research idea needs to be translated into a concrete technical architecture design.
1SKILL.mdUpdated Apr 17, 2026