القائمة
Guide for creating MCP (Model Context Protocol) servers in Python (FastMCP) or Node/TypeScript.
التثبيت باستخدام Codex أو Claude انسخ هذا Prompt والصقه في Codex أو Claude أو مساعد آخر ليراجع صفحة Skill ويثبّتها لك.
استنادا إلى تصنيف SOC المهني
| name | mcp-builder |
| description | Guide for creating MCP (Model Context Protocol) servers in Python (FastMCP) or Node/TypeScript. |
| disable-model-invocation | true |
Proxy to Anthropic's official mcp-builder skill. Read and follow all instructions from the vendor copy:
Read this file first: .claude/vendor/anthropic-skills/skills/mcp-builder/SKILL.md
All supporting files (scripts, references, assets) are in .claude/vendor/anthropic-skills/skills/mcp-builder/.
$ARGUMENTS
Mine architecture decision records (ADRs), gotcha files, runbooks, and docs/archive for cross-project lessons that warrant ai-playbook canonical specs. Use when a project lacks a populated retros/ directory but has equivalent material in other forms.
Implement tasks from an OpenSpec change. Use when the user wants to start implementing, continue implementation, or work through tasks.
Archive a completed change in the experimental workflow. Use when the user wants to finalize and archive a change after implementation is complete.
Orchestrate the canonical task↔PR↔release flow end-to-end. Two modes — `/dev-flow start <description>` scaffolds an OpenSpec change + branch + (optional) worktree + git hook ready for commits; `/dev-flow ship` validates the current branch + pushes + opens PR + monitors CI. Use whenever you start any non-trivial change in a playbook-consuming project. LLM-agnostic — works for Claude Code, Cursor, Antigravity, Gemini CLI, OpenCode, and humans.
Implement an OpenSpec change in parallel using multiple subagents (one per disjoint task group inside the same slice). Use when the slice's tasks.md has ≥2 task groups with disjoint write-paths AND >30 min of parallelisable work. For sequential implementation use openspec-apply-change instead.
Review code changes adversarially using parallel review layers (Blind Hunter, Edge Case Hunter, Acceptance Auditor) with structured triage into actionable categories. Use when the user says "run code review" or "review this code"
