Claude Codepasses structure
Checker 0.1.0 · agent-skills-2026-08-13+claude-code-docs-2026-08-13 · 8/27/2026.claude/skills/ai-eval-toolkitRuntime, accounts, dependencies, permissions, network behavior and task quality remain untested.
Real Skill packageSource verifiedClawHub registry
LLM Evaluation Toolkit（LLM评测工具链）
LLM 评测工具链（可运行实现）——把评测方法论变成能直接跑的本地引擎：评测集管理（JSONL 建集/质量检查/规模统计）、幻觉检测引擎（数字一致性/引用校验/否定矛盾/关键论断互证四类规则检测）、RAG 指标计算（RAGAS 四指标的本地简化实现：忠实度/答案相关性/上下文精度/上下文召回）、回归对比（基线 vs 新结果差异判定）、报告生成与上线门禁（分场景得分/门禁判定/报告输出）。零依赖纯标准库，本地闭环不联网。与「LLM 质量评测」（方法论）互补——那个讲怎么做，这个给能跑的实现。面向 AI 工程师、测试与质量负责人。
Identity and source
Publisher attributionzhaoxinghua09-cellregistry owner unverified by skillvetai
Functional categoryAgent Engineering, Security & Governanceautomatically inferred · 54% rule confidence
Package forminstruction with code19 recorded files
Canonical sourceClawHub registryclawhub:zhaoxinghua09-cell:ai-eval-toolkit
Platform declarations
These states come from the source or distribution context. None of the entries below are SkillVetAI compatibility test results.
OpenClawnative officialProvenance: registry distribution
Independent structural checks
These checks parse the fixed package against dated platform rules. They do not execute the Skill or verify task behavior.
OpenAI Codexpasses structure
Checker 0.1.0 · agent-skills-2026-08-13+codex-docs-2026-08-13 · 8/27/2026.agents/skills/ai-eval-toolkitRuntime, accounts, dependencies, permissions, network behavior and task quality remain untested.
OpenClawpasses structure
Checker 0.1.0 · agent-skills-2026-08-13+openclaw-docs-2026-08-13 · 8/27/2026skills/ai-eval-toolkitRuntime, accounts, dependencies, permissions, network behavior and task quality remain untested.
Installation and inspection
This command is recorded from the source ecosystem and resolves the registry's latest release. The fixed release shown on this page should be inspected before adoption.
clawhub install @zhaoxinghua09-cell/ai-eval-toolkitclawhub inspect @zhaoxinghua09-cell/ai-eval-toolkit --version 1.0.0Security evidence
SkillVetAI static result: no findings detected
This automated, non-executing scan is bound to this release hash. It is not a safety certification and may contain false positives or false negatives.
- Status
- completed
- Coverage
- full text content
- Files
- 18 / 19 inspected as text
- Checked
- 8/27/2026, 10:18:59 PM
- Scanner
- 0.1.3
- Policy
- 1.0.3
3 inferred permission indicators
- shell execution — automatically inferred
- network access — automatically inferred
- filesystem write — automatically inferred
2 dependency and API indicators
- api: clawhub.ai
- api: www.w3.org
External clawhub result: clean
This is registry-supplied evidence for the recorded release, not an independent SkillVetAI scan. Check the canonical source for the full report, scanner versions, scope, and current moderation state.
- Evidence checked
- 8/27/2026, 5:51:09 PM
- Release binding
- Matches this record
- vt: clean
- skillspector: suspicious
- llm: clean
Recorded files
The catalog stores hashes and an inventory summary for change detection. It does not republish the package contents.
Package content hash
sha256:dea7c7d9fb7c4c823a3903e036cc16ac8217de359a905dfbdc6749f50cfd8ce3Show up to 19 recorded paths
ATTESTATION.mdLICENSE.mdmanifest.jsonreferences/01-工具链全景.mdreferences/02-评测集管理.mdreferences/03-幻觉检测引擎.mdreferences/04-RAG指标计算.mdreferences/05-回归对比.mdreferences/06-报告与门禁.mdreferences/07-与平台工具衔接.mdreferences/08-FAQ.mdSECURITY_AUDIT.mdskill-card.mdSKILL.mdtools/eval_toolkit.pyverify/eval_toolkit_quality_test.pyverify/gen_security_radar.pyverify/security_results.jsonverify/security-radar.svg
Source changelog
v1.0.0: runnable LLM evaluation toolkit (zero-dependency, pure stdlib, local offline) - dataset management engine (JSONL init/quality check/statistics), hallucination detection engine (four rule-based detectors, numeric consistency, citation validation, negation contradiction, key-claim cross-checking), RAG metric engine (local simplified RAGAS, faithfulness/answer relevance/context precision/context recall), regression comparison engine (baseline vs new diff), report and launch-gate engine (per-scenario scores, gate decision); complements the LLM quality evaluation methodology playbook