ai-guard-judge-eval
Skill3 starsChanged 26 days agoApache-2.0View on GitHub
USE FOR: benchmarking LLM judges used as safety guards or escalation adjudicators: building a stratified band set from real. DO NOT USE FOR: continuous-score quality rating or preference ranking of chat models. TRIGGERS: LLM judge, safety guard eval, escalation adjudicator.
In this repository65 other files it gives its agents.
- AGENTS.mdAGENTS.md
- applied-data-information-securitySkill
- applied-engineering-disciplinesSkill
- applied-legal-reasoningSkill
- epistemic-discovery-engineSkill
- epistemic-mappingSkill
- hol-metamath-adapterSkill
- isabelle-isar-adapterSkill
- itp-cross-portingSkill
- lean-ai-formalizationSkill
- lean-applied-reasoningSkill
- lean-axiom-closureSkill
- lean-bigraph-modelingSkill
- lean-bisectSkill
- lean-blueprintSkill
- lean-buildSkill
- lean-causal-reasoningSkill
- lean-competitive-mathSkill
- lean-convergence-verificationSkill
- lean-core-onlySkill
- lean-dag-visualizerSkill
- lean-deep-auditSkill
- lean-doc-feedbackSkill
- lean-doc-improvementSkill
- lean-doc-requirementsSkill
- lean-enforcementSkill
- lean-formalization-breakdownSkill
- lean-gatewaySkill
- lean-integration-protocolSkill
- lean-knowledge-formalizationSkill
- lean-math-analysisSkill
- lean-math-discreteSkill
- lean-math-dynamicalSkill
- lean-math-foundationsSkill
- lean-math-optimizationSkill
- lean-math-stochasticSkill
- lean-mweSkill
- lean-nested-learningSkill
- lean-package-researchSkill
- lean-pedagogical-expositionSkill
- lean-proof-refactorSkill
- lean-proof-repairSkill
- lean-proof-reviewSkill
- lean-proofSkill
- lean-prSkill
- lean-quality-engineSkill
- lean-replay-verificationSkill
- lean-reportSkill
- lean-researchSkill
- lean-retroactive-auditSkill
- lean-retro-methodologySkill
- lean-review-councilSkill
- lean-security-formalizationSkill
- lean-setupSkill
- lean-specificationSkill
- lean-tautology-triageSkill
- lean-upstream-portingSkill
- lean-vendor-substrateSkill
- lean-zettelkastenSkill
- math-product-managementSkill
- The first 60 shown.
The file
We split this skill into the rules it gives an agent
4 rules. About 0 are common, 4 are distinctive, and 0 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈4
Distinctive · estimate
This author's own choices. Worth a look.
0
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 24
Run the deterministic gate + active profile over it.
- Distinctive: line 26
Write synthetic probes covering each dangerous-verb family the gate defends (reverse shells, encoded pipes, credential exfil, persistence, …
- Distinctive: line 37
If a judge emits unparseable output during the 3-item smoke check, stop the full run and fix the per-judge schema adapter or disable …
- Distinctive: line 42
If E-band disagreements trace to deterministic gate regex false positives, record them as label noise and fix the gate pattern separately …
≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
ai-guard-judge-eval
Operational dispatch contract for @ai-guard-judge-eval.
Routing
- USE FOR: tasks routed to
@ai-guard-judge-eval. - DO NOT USE FOR: out-of-scope tasks (delegate via
@lean-gateway).
Workflow
- Extract the real command/decision corpus; dedup by exact text.
- Run the deterministic gate + active profile over it; bucket by effective verdict; sample caps per class token for W.
- Write synthetic probes covering each dangerous-verb family the gate defends (reverse shells, encoded pipes, credential exfil, persistence, destruction).
- Emit the set with seed + label provenance in metadata; commit it.
- Per judge: schema adapter + prompt style; smoke on 3 unsafe items (verify parseability) before any full run.
- Full band runs; aggregate per band; record the table.
- For rubric changes: gen N+1 runs full set x repeat 3; adopt only if target band improves AND benign specificity holds within one flip; register the distribution deltas and any negative results.
Recovery & STOP
- If a judge emits unparseable output during the 3-item smoke check, stop the full run and fix the per-judge schema adapter or disable thinking-channel capture before spending full-set compute.
- If E-band disagreements trace to deterministic gate regex false positives, record them as label noise and fix the gate pattern separately rather than tuning the judge rubric to mimic gate bugs.
Handoffs
- Predecessors:
@applied-data-information-security(policy threat model),@lean-ai-formalization(assurance requirements). - Successors:
@lean-ai-formalization(certification evidence),@applied-engineering-disciplines(reliability & testing integration).
Detailed reference
Full encyclopedic reference, decision trees, and lookup tables live in
references/ai-guard-judge-eval-handbook.md .
See also
../../references/ai-guard-judge-eval-handbook.md-- Completeai-guard-judge-evalHandbook../lean-gateway/SKILL.md-- Top-level skill router
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.