用 Codex 或 Claude 帮你安装 复制这段 Prompt，粘贴到 Codex、Claude 或其他助手里，让它检查 Skill 页面并帮你完成安装。
直接命令不会经过审查 Prompt；运行前请先检查来源。
npx skills add https://github.com/Miosa-osa/canopy --skill react-pattern命令会保持在同一行。复制前请横向滚动并检查完整内容。
想先保存到本地？可下载 SkillsMP 当前能够提供的文件。
Evaluate retrieval and generation quality in RAG pipelines. Separate scoring for retrieval (recall, precision, MRR) and generation (faithfulness, relevance, completeness). End-to-end pipeline assessment with bottleneck identification. Triggers on: "eval rag", "rag evaluation", "retrieval evaluation", "rag quality", "rag metrics"
Design binary pass/fail LLM-as-Judge evaluators. Structured prompt engineering for evaluation: criteria definition, rubric construction, few-shot calibration, and bias mitigation. Produces a ready-to-deploy judge prompt with scoring instructions. Triggers on: "judge prompt", "llm judge", "evaluator prompt", "scoring prompt", "grading rubric"
Language Agent Tree Search - Monte Carlo planning - 92.7% on HumanEval
基于 SOC 职业分类
正在显示 SKILL.md
| name | react-pattern |
| description | Thought-Action-Observation loop for transparent reasoning |
| trigger | always |
| priority | 3 |
| dynamic | false |
| created | 2026-01-26 |
Thought → Action → Observation loop for transparent reasoning and reduced hallucination.
Activates on:
THOUGHT: What do I need to do next?
ACTION: [tool_name] with [parameters]
OBSERVATION: [result of action]
... repeat until task complete ...
THOUGHT: Task complete because [reasoning]
Thought: Explicit reasoning about next step
Action: Execute one tool call
Observation: Analyze the result
Works with:
Avoid:
Based on ReAct: Synergizing Reasoning and Acting in Language Models