Agent测试
- 作者仓库星标 2,235
- 作者更新于 实时读取
- 作者仓库 claude-code-plugins-plus-skills
- 领域
- 数据
- 兼容 Agent
-
- Claude Code
- Cursor
- Cline
- Codex
- Windsurf
- Gemini CLI
- +20
- 信任分
- 88 / 100 · 社区维护
- 作者 / 版本 / 许可
- @jeremylongshore · 未声明 license
- Token 消耗评级
- 低消耗
- 接入复杂程度
- 需简单配置
- 是否需要外部 API Key
- 不需要
- 兼容的系统
- Docker
- 底层运行要求
- Docker
- 文件与系统权限
-
- 只读
- 允许写入 / 修改
- 网络行为
- 仅限本地
- 安装命令数
- 26 条
档案由构建时根据 SKILL.md 与安装命令自动衍生,可能与作者实际意图存在差异。
需要注意: 未限定 allowed-tools,默认拥有全部工具权限。
---
name: skill-adapter
description: 'Execute analyzes existing plugins to extract their capabilities, then Analyzes existing plugins…
category: 数据
runtime: Docker
---
# skill-adapter 输出预览
## PART A: 任务判断
- 适用问题:表格、CSV、数据集、指标或分析流程。
- 输入要求:目标材料、限制条件、期望输出和验收方式。
- 证据边界:围绕“Overview / Prerequisites / Instructions”读取原文规则,不把推断写成作者承诺。
## PART B: 执行结果
- **01** 任务判断:确认你的需求是否属于表格、CSV、数据集、指标或分析流程,并标出输入、限制和预期结果。
- **02** 执行计划:优先按“Overview / Prerequisites / Instructions”拆成步骤,说明每一步会读取什么、修改什么、产出什么。
- **03** 交付结果:给出可复制的命令、文件改动、检查清单或内容草稿,并说明如何继续迭代。
- **04** 风险边界:结合 读取文件、写入/修改文件、主要在本地完成、通常不需要额外 API Key 给出执行前确认项。
## Running Rules
- 读取文件、写入/修改文件;主要在本地完成;通常不需要额外 API Key。
- 先小样例验证,再放大到真实任务。
- 交付时同时给结果、检查口径和下一步迭代建议。 原文没有稳定的斜杠命令要求。安装验证后通常全局生效,直接在对话里点名这个 Skill 并描述任务即可。
告诉 Agent 目标文件或材料、期望结果、不可改范围、是否允许联网或执行命令。本 Skill 的权限画像是:读取文件、写入/修改文件。
先用一个小任务确认它会围绕“Overview / Prerequisites / Instructions”工作;涉及文件或命令时,先看 diff、日志、预览或测试结果。
检查最终产物是否包含明确结果、必要证据和下一步动作;如果输出泛泛而谈,就补充输入、边界和验收标准后重跑。
---
name: skill-adapter
description: 'Execute analyzes existing plugins to extract their capabilities, then Analyzes existing plugins…
category: 数据
source: jeremylongshore/claude-code-plugins-plus-skills
---
# skill-adapter
## 什么时候使用
- 把数据处理方向的常用动作沉淀成 Agent 可调用的技能 适合处理表格、CSV、指标、数据集、分析和可视化报告,核心价值是把输入、判断、执行、验证和交付边界固定下来,避免 Agent 泛泛回答。 把任务拆成可执行、可检查、可继续迭代的步…
- 面向表格、CSV、数据集、指标或分析流程,优先处理能明确输入、步骤和验收标准的工作。
## 需要提供什么
- 目标材料、目录范围、期望结果和不可改动内容。
- 是否允许联网、执行命令、读写文件或调用外部服务。
## 执行规则
- 围绕「Overview / Prerequisites / Instructions」组织步骤,不把推断写成作者事实。
- 读取文件、写入/修改文件;主要在本地完成;通常不需要额外 API Key。
- 先跑小样例,确认结果可检查后再扩大任务范围。
## 输出要求
- 给出最终产物、关键证据、验证方式和下一步动作。
- 信息不足时标记 unknown,不编造命令、平台或依赖。 作者原文负责流程事实;仓库文件负责来源和命令;流狐只补充适用场景、限制和质量判断。
skill "skill-adapter" {
输入层 -> 用户目标 + 目标文件 + 禁止范围 + 验收标准
上下文层 -> Overview / Prerequisites / Instructions
规则层 -> SKILL.md 触发条件 / 执行顺序 / 输出格式
运行层 -> Docker | 读取文件、写入/修改文件 | 主要在本地完成
安全层 -> 通常不需要额外 API Key + 小任务验证 + diff / 日志复核
输出层 -> 可复制结果 + 检查清单 + 下一步迭代
} Skill Adapter
Overview
Analyzes existing plugins in the repository to extract their capabilities, then synthesizes and applies those learned patterns to the current task. Functions as a universal skill chameleon that discovers relevant plugins, extracts their approaches and methodologies, and adapts them to novel requests.
Prerequisites
- Read access to the
plugins/directory tree (community, packages, examples categories) grepandfindavailable on PATH for plugin discovery- Familiarity with the plugin structure:
commands/*.md,agents/*.md,skills/*/SKILL.md, andscripts/
Instructions
- Analyze the user's task to identify the core capability needed, the domain (security, devops, testing, documentation, etc.), and key requirements or constraints (see
${CLAUDE_SKILL_DIR}/references/how-it-works.md). - Search existing plugins for relevant capabilities using file globbing across
plugins/community/,plugins/packages/, andplugins/examples/directories. Match onplugin.jsondescriptions and keyword fields. - For each relevant plugin discovered, extract capabilities from its components:
- Commands (
commands/*.md): read content, extract approach and input/output patterns. - Agents (
agents/*.md): understand roles, decision-making patterns, expertise areas. - Skills (
skills/*/SKILL.md): read instructions, extract core capability and tool usage. - Scripts (
scripts/*.sh,*.py): analyze logic, identify reusable patterns and error handling.
- Commands (
- Synthesize extracted patterns by merging complementary approaches, simplifying where possible, and ensuring compatibility with the current environment.
- Apply the adapted skill to the user's task, following the learned methodology while adjusting syntax, tools, and output format to match the current context.
- Report which plugins were consulted, what patterns were extracted, and how they were adapted for the current task.
Output
A structured adaptation report containing:
- List of plugins analyzed and capabilities extracted from each
- The synthesized approach combining relevant patterns
- The direct application of that approach to the user's task
- Any caveats or limitations of the adapted skill
Error Handling
| Error | Cause | Solution |
|---|---|---|
| No matching plugins found | Search terms too narrow or domain not represented | Broaden search keywords; check alternative categories; fall back to general-purpose approach |
| Plugin directory inaccessible | Missing read permissions or incorrect path | Verify plugins/ directory exists and permissions allow traversal |
| Incompatible patterns | Extracted approaches conflict with current environment | Prioritize the most relevant plugin's approach; discard conflicting elements |
| Empty skill/command files | Plugin has stub content without real instructions | Skip that plugin and note it as incomplete; rely on other sources |
Examples
Learning code analysis from security plugins:
Task: "Analyze this codebase for issues."
Process: Discover owasp-top-10-scanner, code-quality-enforcer, and security-audit-agent. Extract OWASP vulnerability checks, complexity/duplication metrics, and dependency scanning patterns. Synthesize a multi-layer analysis covering security, quality, and dependencies. Apply to the target codebase (see ${CLAUDE_SKILL_DIR}/references/example-workflows.md).
Adopting documentation skills:
Task: "Generate API documentation."
Process: Find api-documenter, openapi-generator, readme-builder. Extract code parsing, OpenAPI spec generation, and hierarchical documentation structuring. Combine into an end-to-end pipeline: parse endpoints, generate spec, create interactive docs, build README.
Learning automation from DevOps plugins: Task: "Automate deployment process." Process: Search DevOps category for deployment, CI/CD, and Docker plugins. Extract build-test-deploy-verify workflows, parallel job patterns, and service orchestration. Adapt to the user's specific tech stack and infrastructure.
Resources
${CLAUDE_SKILL_DIR}/references/how-it-works.md-- detailed five-phase adaptation process${CLAUDE_SKILL_DIR}/references/example-workflows.md-- end-to-end workflow examples${CLAUDE_SKILL_DIR}/references/errors.md-- error handling patterns
先判断是否适合
作者设计意图
作者的方法与取舍
边界和复核