Agent安装
- 作者仓库星标 39
- 作者更新于 实时读取
- 作者仓库 awesome-omni-skill
- 领域
- AI 智能
- 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
- @diegosouzapw · 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
- 未声明(默认跨平台)
- 底层运行要求
- Python
- 文件与系统权限
-
- 只读
- 允许写入 / 修改
- Shell 执行
- 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: setup-asdlc
description: Initialize a repository for ASDLC adoption with AGENTS.md and directory structure Enable teams t…
category: AI 智能
runtime: Python
---
# setup-asdlc 输出预览
## PART A: 任务判断
- 适用问题:提示词、Agent 工作流、模型评估或自动化推理。
- 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 / Definitions / Prerequisites”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提示词、Agent 工作流、模型评估或自动化推理,并标出输入、限制和预期结果。
- **02** 执行计划:优先按“Overview / Definitions / Prerequisites”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、执行终端命令、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、执行终端命令;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。 原文出现了 `/create-plan`、`/setup-asdlc` 这类斜杠命令;如果你的 Agent 支持命令触发,优先用命令开场,再补充目标和边界。
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、执行终端命令。
先用一个小任务确认它会围绕“Overview / Definitions / Prerequisites”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: setup-asdlc
description: Initialize a repository for ASDLC adoption with AGENTS.md and directory structure Enable teams t…
category: AI 智能
source: diegosouzapw/awesome-omni-skill
---
# setup-asdlc
## 什么时候使用
- 把 AI / Agent方向的常用动作沉淀成 Agent 可调用的技能 适合处理AI Agent、提示词、模型评估与自动化推理,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…
- 面向提示词、Agent 工作流、模型评估或自动化推理,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 / Definitions / Prerequisites」组织步骤,不把推断写成作者事实。
- 读取文件、写入/修改文件、执行终端命令;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 "setup-asdlc" {
输入层 -> 用户目标 + 目标文件 + 禁止范围 + 验收标准
上下文层 -> Overview / Definitions / Prerequisites
规则层 -> SKILL.md 触发条件 / 执行顺序 / 输出格式
运行层 -> Python | 读取文件、写入/修改文件、执行终端命令 | 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
} Setup ASDLC
Overview
Initialize a repository for ASDLC adoption by creating AGENTS.md template, directory structure (specs/, .plans/), and basic configuration. This command prepares repositories to use ASDLC patterns while remaining optional—other commands work without running setup.
Definitions
- AGENTS.md: Agent Constitution file that defines project context, operational boundaries, and command registry for AI agents.
- specs/: Directory containing permanent living specifications for features.
- .plans/: Directory containing transient task-level implementation plans.
- schemas/: Directory containing JSON Schema definitions for validation (Standardized Parts pattern).
- MCP: Model Context Protocol — framework for AI agents to connect to external services.
- Standardized Parts: ASDLC pattern for schema-enforced structure and validation.
Prerequisites
- None required. This command is optional and can be run at any time.
- MCP setup is optional — command will check MCP status but won't block if unavailable.
- Git repository — command works best in a Git repository but doesn't require it.
Purpose
Enable teams to quickly adopt ASDLC patterns by automating initial repository setup. This command creates the foundational structure (AGENTS.md, specs/, .plans/) that supports ASDLC workflows while maintaining flexibility—teams can use other commands without running setup.
Steps
Check for existing AGENTS.md
- Check if
AGENTS.mdexists in repository root - If exists: Report "AGENTS.md already exists, skipping creation"
- If missing: Proceed to generate template (Step 2)
- Check if
Generate AGENTS.md template (if missing)
- Read existing
AGENTS.mdfrom this repository as template reference (if available) - Create
AGENTS.mdwith template structure:- Project Mission placeholder:
> **Project Mission:** [Your project mission statement] - Core Philosophy placeholder:
> **Core Philosophy:** [Your core development philosophy] - ASDLC Alignment placeholder:
> **ASDLC Alignment:** [Your ASDLC alignment statement] - Identity & Persona section with placeholders
- Tech Stack section with placeholders
- Operational Boundaries (3-tier structure):
- Tier 1 (ALWAYS): Non-negotiable standards
- Tier 2 (ASK): High-risk operations requiring Human-in-the-Loop
- Tier 3 (NEVER): Safety limits
- Command Registry (empty, to be populated)
- Development Map (placeholder)
- Common Pitfalls section (examples)
- Project Mission placeholder:
- Report: "Created AGENTS.md template"
- Read existing
Create specs/ directory
- Check if
specs/directory exists - If exists: Report "specs/ directory already exists, skipping creation"
- If missing: Create
specs/directory - Check if
specs/README.mdexists - If missing: Copy
specs/README.mdfrom this repository (if available) or create basic README - Report: "Created specs/ directory with README.md"
- Check if
Create .plans/ directory
- Check if
.plans/directory exists - If exists: Report ".plans/ directory already exists, skipping creation"
- If missing: Create
.plans/directory - Create
.plans/.gitkeepfile to ensure directory is tracked by Git - Report: "Created .plans/ directory"
- Check if
Detect if schemas/ directory is needed
- Check if
schemas/directory exists - If exists: Report "schemas/ directory already exists, skipping creation"
- If missing: Check for schema indicators:
- If
AGENTS.mdexists, search for keywords: "Standardized Parts", "schema", "validation" - If any keyword found: Create
schemas/directory andschemas/README.mdtemplate - Report: "Detected need for schemas/ (found: {keyword} in AGENTS.md)" or "No schemas/ needed (no indicators found)"
- If
- Check if
Create schemas/README.md template (if schemas/ was created)
- Create
schemas/README.mdwith template explaining:- When schemas are needed (Standardized Parts pattern)
- Link to ASDLC Standardized Parts pattern
- Note: Can be added later if needed
- Report: "Created schemas/README.md template"
- Create
Optional: Verify MCP setup (non-blocking)
- Attempt to run
python schemas/validate_mcps.py --list(if available) - If successful: Report "MCP setup detected"
- If fails or unavailable: Report "MCP setup not detected (optional, can be configured later)"
- Continue regardless of result
- Attempt to run
Optional: Verify issue tracker connection (non-blocking)
- Attempt to call lightweight MCP tool (e.g.,
mcp_atlassian_getAccessibleAtlassianResourcesormcp_github_list_commits) - If successful: Report "Issue tracker connection verified"
- If fails or unavailable: Report "Issue tracker connection not available (optional, can be configured later)"
- Continue regardless of result
- Attempt to call lightweight MCP tool (e.g.,
Generate setup summary
- Report what was created vs skipped
- Provide next steps guidance:
- "Next steps: Customize AGENTS.md with your project details"
- "Configure MCP servers (see docs/reference/mcp-setup.md)"
- "Create your first spec with /create-plan"
Tools
Filesystem Tools
read_file- Read existing AGENTS.md and specs/README.md as templateswrite- Create new files (AGENTS.md, specs/README.md, schemas/README.md)list_dir- Check if directories existglob_file_search- Find existing files
Terminal Tools
run_terminal_cmd- Execute commands:python schemas/validate_mcps.py --list(if available) - Check MCP setupgit status(optional) - Verify Git repository
MCP Tools (Optional, Non-Blocking)
mcp_atlassian_getAccessibleAtlassianResources- Verify Atlassian/Jira connection- Parameters: None
- Error handling: If fails, report but continue
mcp_github_list_commits- Verify GitHub connection- Parameters:
owner,repo(if available) - Error handling: If fails, report but continue
- Parameters:
Expected Output
Fresh Repository Setup
🚀 ASDLC Setup Complete
Created:
✅ AGENTS.md template
✅ specs/ directory with README.md
✅ .plans/ directory
⚠️ schemas/ not needed (no indicators found)
MCP Status:
⚠️ MCP setup not detected (optional, can be configured later)
⚠️ Issue tracker connection not available (optional, can be configured later)
Next steps:
1. Customize AGENTS.md with your project details
2. Configure MCP servers (see docs/reference/mcp-setup.md)
3. Create your first spec with /create-plan
Partial Setup (AGENTS.md exists)
🚀 ASDLC Setup Complete
Skipped:
⚠️ AGENTS.md already exists
Created:
✅ specs/ directory with README.md
✅ .plans/ directory
✅ schemas/ directory (detected: Standardized Parts in AGENTS.md)
MCP Status:
✅ MCP setup detected
✅ Issue tracker connection verified
Next steps:
1. Create your first spec with /create-plan
Full Setup Already Exists
🚀 ASDLC Setup Complete
All directories and files already exist:
✅ AGENTS.md
✅ specs/ directory
✅ .plans/ directory
✅ schemas/ directory
MCP Status:
✅ MCP setup detected
✅ Issue tracker connection verified
Repository is already set up for ASDLC!
When to Use
- New repository - Set up ASDLC structure from scratch
- Existing repository - Add ASDLC structure to existing project
- Team onboarding - Standardize setup across team members
- ASDLC adoption - Begin using ASDLC patterns and workflows
Error Handling
If AGENTS.md template generation fails:
- Report error but continue with directory creation
- Provide manual creation instructions
If directory creation fails:
- Report which directories failed
- Provide manual creation instructions
- Continue with other operations
If MCP verification fails:
- Report warning but continue (MCP is optional)
- Provide MCP setup documentation link
If file copy fails (specs/README.md):
- Create basic README.md with minimal content
- Report that template copy failed
Notes
- This command is optional — other commands work without running setup
- Command never overwrites existing files or directories
- MCP verification is non-blocking — command succeeds even if MCP is unavailable
- Schema detection is intelligent — only creates schemas/ if indicators are found
- All operations are idempotent — safe to run multiple times
Guidance
Role
Act as a setup assistant helping teams initialize their repository for ASDLC adoption. You are helpful, non-intrusive, and provide clear guidance.
Instruction
Execute the setup workflow to prepare a repository for ASDLC by creating AGENTS.md template, directory structure, and optional configuration checks. Always check for existing files before creating, never overwrite, and provide clear feedback on what was created vs skipped.
Context
- This command enables teams to quickly adopt ASDLC patterns
- Command is optional—other commands work without running setup
- Based on recommendations from FB-32 analysis
- Follows skill structure standards from AGENTS.md
- ASDLC patterns: Agent Constitution, The Spec, Standardized Parts
- ASDLC pillars: Factory Architecture (command station), Standardized Parts (schema-enforced structure), Quality Control (validation gates)
Examples
ASDLC: Agent Constitution — AGENTS.md implements this pattern. Standardized Parts — Schema detection aligns with this pattern.
Example: Fresh Repository
User runs: /setup-asdlc
Result: Creates AGENTS.md, specs/, .plans/, reports summary
Example: Existing Repository
User runs: /setup-asdlc
Result: Skips existing files, creates missing directories, reports what was added
Constraints
Rules (Must Follow):
- Operational Standards Compliance: This command follows operational standards (documented in AGENTS.md if present, but apply universally):
- File Operations: Follow best practices for file operations and directory structure
- Safety Limits: Never commit secrets, API keys, or sensitive data in generated files
- AGENTS.md Optional: This command creates AGENTS.md, but other commands work without it. Standards apply regardless.
- See AGENTS.md §3 Operational Boundaries (if present) for detailed standards
- Never Overwrite - Always check for existing files/directories before creating
- Optional MCP - MCP verification must not block command execution
- Intelligent Detection - Only create schemas/ if indicators are found in AGENTS.md
- Clear Feedback - Report what was created vs skipped with clear messages
- Idempotent - Command must be safe to run multiple times
- Template-Based - Use existing AGENTS.md and specs/README.md as templates when available
Existing Standards (Reference):
- Command structure: See AGENTS.md §6 Command Structure Standards
- Directory conventions: specs/ for specs, .plans/ for plans
- File naming: AGENTS.md at root, specs/README.md for specs documentation
- MCP setup: See docs/reference/mcp-setup.md for MCP configuration
- Schema validation: See schemas/README.md for when schemas are needed
Output
- Setup Summary: Report of what was created vs skipped
- MCP Status: Optional verification results (non-blocking)
- Next Steps: Guidance on customizing AGENTS.md and configuring MCP
- Clear Feedback: Explicit messages for each operation (created, skipped, detected)
The command prepares the repository for ASDLC adoption while maintaining flexibility and never overwriting existing files.
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