{
  "markdown": "# DevPlan MCP Server\n\n[![License: MIT](https://img.shields.io/badge/License-MIT-yellow.svg)](https://opensource.org/licenses/MIT)\n[![MCP](https://img.shields.io/badge/MCP-Model%20Context%20Protocol-purple.svg)](https://modelcontextprotocol.io)\n[![Cloudflare Workers](https://img.shields.io/badge/Cloudflare-Workers-orange.svg)](https://workers.cloudflare.com/)\n[![21 Tools](https://img.shields.io/badge/Tools-21-blue.svg)](#tools)\n\n**Transform ideas into executable development plans** — an MCP server that brings the [ClaudeCode-DevPlanBuilder](https://github.com/mmorris35/ClaudeCode-DevPlanBuilder) methodology to Claude Code.\n\n> **The Problem**: AI coding assistants often lose context, skip steps, or produce inconsistent code across sessions.\n>\n> **The Solution**: DevPlan creates detailed, agent-executable development plans with built-in validation, lessons learned, and inline git workflows.\n\n---\n\n## NEW — Install as a Skill (No MCP Server Required)\n\nDevPlan is now available as a **standalone Claude Code skill** — no MCP server, no network dependency, no SSE timeouts. The entire DevPlan methodology runs locally as a first-class plugin.\n\n### Install\n\n**Step 1** — Add this repo as a plugin marketplace:\n\n```bash\n/plugin marketplace add mmorris35/devplan-mcp-server\n```\n\n**Step 2** — Install the plugin at user scope (available across all projects):\n\n```bash\n/plugin install devplan@mmorris35 --scope user\n```\n\n**Step 3** — Reload plugins to activate:\n\n```bash\n/reload-plugins\n```\n\n### Usage\n\nInvoke the skill directly:\n\n```\n/devplan\n```\n\nOr use specific sub-commands:\n\n```\n/devplan brief          — Create or parse a PROJECT_BRIEF.md\n/devplan plan           — Generate a Haiku-executable DEVELOPMENT_PLAN.md\n/devplan agents         — Generate executor (Haiku) and verifier (Sonnet) agents\n/devplan claude-md      — Generate a project CLAUDE.md\n/devplan validate       — Check plan structure and Haiku-executability\n/devplan progress       — Show completion status\n/devplan export mermaid — Generate a Mermaid flowchart of the plan\n/devplan issue <number> — Convert a GitHub issue to a remediation task\n/devplan implement      — Kickoff the build with executor + verifier agents\n```\n\nOr just describe what you need — Claude will invoke the skill automatically:\n\n```\n\"Help me plan a CLI tool for managing dotfiles\"\n\"Create a development plan for this project\"\n\"Validate my development plan\"\n```\n\n### What's Included\n\n```\nskills/devplan/\n├── SKILL.md                 — Core methodology, interview flow, dispatch\n├── references/\n│   ├── templates.md         — Brief/plan/CLAUDE.md templates (CLI, web app, API, library)\n│   ├── validation.md        — Structure + Haiku-executability rules + battle-tested lessons\n│   ├── agents.md            — Executor and verifier agent generation patterns\n│   └── workflows.md         — Mermaid/ReactFlow export + progress tracking\n├── scripts/\n│   ├── validate-plan.sh     — Structural validation (standalone)\n│   └── check-haiku.sh       — Haiku-executability checker (standalone)\n└── examples/\n    └── hello-cli-plan.md    — Gold standard reference plan\n```\n\n### Skill vs MCP Server\n\n| | Skill (NEW) | MCP Server |\n|---|---|---|\n| **Network** | None required | SSE connection to Cloudflare |\n| **Reliability** | Always works | Subject to SSE timeouts |\n| **Lessons system** | Use [Nellie](https://github.com/mmorris35/nellie) or your own | Built-in KV store |\n| **Install** | `/plugin install` | `claude mcp add` |\n| **Validation scripts** | Standalone bash | Server-side |\n\nBoth options are fully supported. The skill is recommended for reliability; the MCP server adds the lessons learned system and usage analytics.\n\n---\n\n## Key Features\n\n| Feature | Description |\n|---------|-------------|\n| **Agent-Executable Plans** | Plans so detailed that any LLM coding agent can execute them mechanically |\n| **Built-in Validation** | Validates plans are complete before execution begins |\n| **Real-Time Progress Tracking** | Integrates with Claude Code's Task tools for live visibility |\n| **Lessons Learned** | Captures issues from verification and injects them into future plans |\n| **Issue Remediation** | Converts GitHub issues directly into remediation tasks |\n| **Executor & Verifier Agents** | Auto-generates specialized agents with task tracking built-in |\n\n## Install\n\n```bash\nclaude mcp add devplan --transport sse https://mcp.devplanmcp.store/sse --scope user\n```\n\nOr add to `~/.claude.json` under the `mcpServers` key:\n\n```json\n{\n  \"mcpServers\": {\n    \"devplan\": {\n      \"type\": \"sse\",\n      \"url\": \"https://mcp.devplanmcp.store/sse\"\n    }\n  }\n}\n```\n\n### Update Existing Installation\n\nIf you already have DevPlan installed, remove from both scopes and re-add:\n\n```bash\nclaude mcp remove devplan --scope project; claude mcp remove devplan --scope user; claude mcp add devplan --transport sse https://mcp.devplanmcp.store/sse --scope user\n```\n\n## Quick Start\n\n```\nYou: \"Use devplan_start to help me build a CLI tool for managing dotfiles\"\n```\n\nThat's it. DevPlan will guide Claude through the entire process.\n\n## The DevPlan Workflow\n\nDevPlan uses a **scaffold → enhance → validate** workflow that ensures every plan is agent-executable before implementation begins.\n\n```mermaid\nflowchart LR\n    subgraph Planning[\"📋 Planning\"]\n        A[Interview] --> B[Brief]\n        B --> C[Generate Scaffold]\n    end\n\n    subgraph Enhancement[\"✨ Enhancement\"]\n        C --> D[Enhance with Code]\n        D --> E{Validate}\n        E -->|Fail| D\n        E -->|Pass| F[Ready]\n    end\n\n    subgraph Execution[\"⚡ Execution\"]\n        F --> G[Agent Executes]\n        G --> H[Agent Verifies]\n    end\n\n    subgraph Learning[\"🧠 Learning\"]\n        H -->|issues| I[Capture Lessons]\n        I -->|improve| C\n    end\n\n    style E fill:#fff3e0,stroke:#f57c00\n    style F fill:#c8e6c9,stroke:#2e7d32\n    style I fill:#e3f2fd,stroke:#1565c0\n```\n\n### How It Works\n\n1. **Interview** → DevPlan asks questions to understand your project\n2. **Brief** → Creates a structured PROJECT_BRIEF.md with requirements\n3. **Generate Scaffold** → `devplan_generate_plan` creates foundation phases + a feature list\n4. **Enhance with Code** → Your AI agent structures feature phases and fills in complete, copy-pasteable code\n5. **Validate** → `devplan_validate_plan` checks the plan is agent-executable\n6. **Execute** → Executor agent implements each subtask with inline git commands\n7. **Verify** → Verifier agent tries to break the implementation\n8. **Learn** → Issues become lessons for future projects\n\n### Validation Ensures Quality\n\nThe validation step checks that plans are truly executable:\n\n- ✅ Complete code blocks (not pseudocode or placeholders)\n- ✅ All imports included in code blocks\n- ✅ No \"add to existing\" instructions\n- ✅ No cross-subtask references\n- ✅ Verification commands with expected outputs\n\n```\n# Example validation output\n{\n  \"valid\": true,\n  \"errors\": [],\n  \"warnings\": [],\n  \"stats\": {\n    \"subtasks\": 5,\n    \"codeBlocksChecked\": 8,\n    \"issuesFound\": 0\n  }\n}\n```\n\n### Real-Time Progress with Task Tools\n\nGenerated executor and verifier agents integrate with Claude Code's Task tools for live progress visibility:\n\n- **Executor agents** create tasks for each subtask, showing real-time spinners as work progresses\n- **Verifier agents** create tasks for each verification phase (Smoke Tests, Feature Verification, Edge Cases, etc.)\n- Progress is visible without scrolling — you always know what Claude is working on\n\n```\n# Example: Executor tracks subtasks\nTaskCreate({ subject: \"1.2.3: Implement auth middleware\", activeForm: \"Implementing auth middleware\" })\nTaskUpdate({ taskId: \"...\", status: \"in_progress\" })\n# ... work happens ...\nTaskUpdate({ taskId: \"...\", status: \"completed\" })\n```\n\nBoth Task tools (real-time visibility) and DEVELOPMENT_PLAN.md (durable record) are updated — giving you the best of both worlds.\n\n## Usage Examples\n\n### New Project\n```\n\"Use devplan_start to help me build [your idea]\"\n```\n\n### Fix a GitHub Issue\n```bash\n# Get issue JSON\ngh issue view 123 --json number,title,body,labels,comments,url > issue.json\n\n# Then tell Claude:\n\"Use devplan_issue_to_task with this issue to create a remediation plan\"\n```\n\n### Check Progress\n```\n\"Use devplan_progress_summary to show me where we are\"\n```\n\n## Multi-Model Support\n\nDevPlan generates plans and agent files for multiple AI coding tools and models. Use the `target` parameter to generate outputs for your preferred tool.\n\n### Supported Targets\n\n| Target | Tool | Agent File | Best For |\n|--------|------|-----------|----------|\n| `claude` | Claude Code | `CLAUDE.md` | Claude Code IDE (default) |\n| `cursor` | Cursor IDE | `.cursorrules` | Cursor AI editor |\n| `aider` | Aider CLI | `.aider.conf.yml` | Terminal-based AI pair programming |\n| `cline` | VS Code Cline | `.cline/instructions.md` | VS Code extension |\n| `windsurf` | Windsurf IDE | `.windsurf/rules.md` | Codium's AI IDE |\n| `generic` | Any Model | `AGENTS.md` + files | Model-agnostic markdown format |\n\n### Using Targets\n\nWhen generating plans or agent files, specify the target tool:\n\n**Generate plan for Cursor:**\n```\n\"Use devplan_generate_plan to create a plan, then I'll customize it for Cursor. Set target to 'cursor' for .cursorrules format\"\n```\n\n**Generate executor for Aider:**\n```\n\"Use devplan_generate_executor with target='aider' to create an Aider-compatible executor agent\"\n```\n\n**Generate generic agent files:**\n```\n\"Use devplan_generate_claude_md with target='generic' to create model-agnostic AGENTS.md files\"\n```\n\n### How Targets Work\n\nEach target has a dedicated **adapter** that transforms the DevPlan methodology into the appropriate format:\n\n- **Claude** - Generates `CLAUDE.md` with executor/verifier agents in `.claude/agents/`\n- **Cursor** - Generates `.cursorrules` with all guidance in one file (Cursor doesn't support separate agents)\n- **Aider** - Generates `.aider.conf.yml` with architect mode instructions\n- **Cline** - Generates `.cline/instructions.md` with executor/verifier split\n- **Windsurf** - Generates `.windsurf/rules.md` with cascade-optimized format\n- **Generic** - Generates `AGENTS.md`, `EXECUTOR.md`, and `VERIFIER.md` for any tool\n\n### Examples\n\n**Start a new project for Cursor:**\n```\n\"Use devplan_start to help me build a CLI tool, then generate the plan with target='cursor' for Cursor IDE\"\n```\n\n**Add executor for specific target:**\n```\n\"I have a development plan. Use devplan_generate_executor with target='aider' to create the executor agent for Aider\"\n```\n\n**Compare adapter capabilities:**\nSee [docs/ADAPTERS.md](/docs/ADAPTERS.md) for a detailed comparison of each target's capabilities and limitations.\n\n## Tools\n\n### Planning\n\n| Tool | Purpose |\n|------|---------|\n| `devplan_start` | Main entry point - guides Claude through the methodology |\n| `devplan_interview_questions` | Get questions to gather project requirements |\n| `devplan_create_brief` | Generate PROJECT_BRIEF.md |\n| `devplan_parse_brief` | Parse existing brief into structured data |\n| `devplan_list_templates` | List project templates (cli, web_app, api, library) |\n\n### Generation\n\n| Tool | Purpose |\n|------|---------|\n| `devplan_generate_plan` | Generate DEVELOPMENT_PLAN.md scaffold with validation instructions |\n| `devplan_generate_claude_md` | Generate CLAUDE.md scaffold |\n| `devplan_generate_executor` | Generate executor agent with Task tool integration |\n| `devplan_generate_verifier` | Generate verifier agent with Task tool integration |\n\n### Validation & Execution\n\n| Tool | Purpose |\n|------|---------|\n| `devplan_validate_plan` | Validate plan structure and agent-executability |\n| `devplan_get_subtask` | Get specific subtask details by ID |\n| `devplan_update_progress` | Mark subtasks complete with notes |\n| `devplan_progress_summary` | Get completion stats and next actions |\n\n### Lessons Learned\n\nFeedback loop that captures issues from verification and incorporates them into future plans.\n\n| Tool | Purpose |\n|------|---------|\n| `devplan_add_lesson` | Capture a lesson from verifier findings |\n| `devplan_list_lessons` | List accumulated lessons by severity |\n| `devplan_archive_lesson` | Archive old lessons without deleting them |\n| `devplan_delete_lesson` | Remove outdated or incorrect lessons |\n| `devplan_extract_lessons_from_report` | Auto-extract lessons from verification reports |\n\n### Issue Remediation\n\nConvert GitHub issues into structured remediation tasks — perfect for bug fixes and post-release maintenance.\n\n| Tool | Purpose |\n|------|---------|\n| `devplan_parse_issue` | Analyze a GitHub issue to extract requirements |\n| `devplan_issue_to_task` | Generate remediation task with subtasks from an issue |\n\n### Analytics\n\n| Tool | Purpose |\n|------|---------|\n| `devplan_usage_stats` | View usage distribution across users |\n\n## Why DevPlan?\n\n| Without DevPlan | With DevPlan |\n|-----------------|--------------|\n| Context lost between sessions | Plans preserve full context |\n| Inconsistent code quality | Executor agents follow exact specifications |\n| Same mistakes repeated | Lessons learned system prevents recurrence |\n| No verification step | Verifier agents actively try to break the code |\n| Bugs found in production | Issues caught before release |\n| Plans need interpretation | Validated plans are copy-paste ready |\n\n## Dashboard & Analytics\n\nDevPlan includes a public dashboard for viewing aggregate usage statistics:\n\n**Dashboard URL**: [devplanmcp.store/dashboard](https://devplanmcp.store/dashboard)\n\nThe dashboard shows:\n- **Summary cards**: Total sessions, total tool calls, countries reached\n- **Line chart**: Sessions and tool calls over the last 30 days\n- **Country table**: Top 10 countries by session count\n\n### Privacy\n\nAll analytics are privacy-preserving:\n- **No IP storage**: Only Cloudflare-derived country/region codes\n- **No user identification**: Sessions are anonymous\n- **Auto-expiration**: Daily stats expire after 90 days via KV TTL\n\n## Development\n\n```bash\nnpm install\nnpm run dev      # Local development\nnpm run deploy   # Deploy to Cloudflare Workers\n```\n\n## Contributing\n\nContributions welcome! Please see the [ClaudeCode-DevPlanBuilder](https://github.com/mmorris35/ClaudeCode-DevPlanBuilder) repo for methodology details.\n\n## License\n\nMIT\n\n---\n\n<p align=\"center\">\n  <b>Built for Claude Code</b><br>\n  <a href=\"https://modelcontextprotocol.io\">Model Context Protocol</a> •\n  <a href=\"https://workers.cloudflare.com/\">Cloudflare Workers</a> •\n  <a href=\"https://github.com/mmorris35/ClaudeCode-DevPlanBuilder\">DevPlanBuilder Methodology</a>\n</p>\n",
  "bytes": 14532,
  "sha": "db515df06b6829641368145aacc6effd1c54c3a0b556e2c57b6b70d52814edf2",
  "repo_slug": "mmorris35/devplan-mcp-server",
  "fonte": "repo",
  "truncated": false,
  "api": "https://agentalog.com/api/listings/mcp_io_github_mmorris35_devplan_mcp_server_5ef59fd1/readme"
}