The ToolPlan MCP server provides a single tool, plan_project, that transforms a raw project idea into a structured, cost-aware, markdown-formatted prompt for AI coding agents.
What it does:
Takes a raw project idea and quality grade to produce an enriched prompt containing a project brief, recommended tech stack, high-leverage open-source tools/MCPs/skills that AI models typically miss, cost-saving execution directives (e.g., subagent delegation, plan-first execution), and source references.
Leverages a bundled, curated knowledge base (kb/*.yaml) covering stacks, tools, MCPs, skills, and directives — no live network calls at runtime.
Designed to make projects more "one-shotable" by producing a superior first prompt, even on non-frontier models.
Input Parameters:
idea (required, string, min 10 chars) — Your raw project description
grade (required: "industry" or "personal") — The quality/professionalism target
tags (optional, array of strings) — Additional tags to guide knowledge base matching
Key Characteristics:
Runs entirely locally; no external data is sent (privacy-preserving)
Single-shot execution — no long-running or interactive tasks
Optional local logging for usage analysis
Click on "Deploy Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type @ followed by the MCP server name and your instructions, e.g., "@ToolPlan MCP plan a personal habit tracker app"
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
Turn a raw project idea into a polished, cost-aware first prompt for any AI coding agent.
Models pick tech stacks decently — but they:
Under-recommend high-leverage open-source tools. Niche skills, scrapers, and community tools from GitHub/Reddit (Agent-Reach, caveman, karpathy-guidelines, ...) save hours, and models rarely surface them unprompted. Programmers who don't track this ecosystem lose that time.
Never apply cost-saving techniques on their own. Subagent delegation, plan-first execution, task-by-task verification — models don't do these unless told, and casual users don't know to ask. Result: millions of wasted tokens.
Do better with a structured first prompt. A polished brief with stack, constraints, and done-criteria makes a project far more one-shotable — even on non-frontier models.
ToolPlan packages all three into one MCP tool call.
Related MCP server: agentforge-mcp
your raw idea ──▶ plan_project(idea, grade) ──▶ enriched prompt
│
reads curated KB (kb/*.yaml):
stacks · tools · MCPs · skills · directives
each with why_models_miss_it + cost_profileNo live scraping at runtime — a weekly offline pipeline proposes KB updates as human-reviewed diffs, so advice stays current without hype pollution.
claude mcp add toolplan -- npx -y toolplan-mcpThen either:
/toolplan <your idea> — copy commands/toolplan.md (shipped in the
npm package) to ~/.claude/commands/ first. The agent calls the tool,
shows you the refined prompt verbatim, and waits for you to proceed, edit,
or regenerate — it never starts building on its own.
/mcp__toolplan__plan — zero-install; Claude Code auto-exposes the
server's built-in plan prompt as a slash command.
Or just ask in chat: "Use plan_project with my idea: an app that tracks freelance invoices, grade personal."
Other hosts (Cursor, Codex CLI, any stdio MCP host): see docs/HOST_SETUP.md.
plan_project(idea: string, grade: "industry" | "personal", tags?: string[])
→ markdown enriched prompt: project brief, recommended stack, tools you'd
likely miss, execution directives, quality bar, sources.
One YAML file per entry under kb/<category>/. Format: docs/KB_SCHEMA.md.
Contributions welcome — PRs must pass the eval regression suite.
Privacy note: running the tool never phones home. The KB is read-only at
runtime and bundled with the package; nobody's usage updates it. Optional
TOOLPLAN_LOG writes usage lines to a local file you control.
Three ways, smallest first:
Add one entry by hand. Copy an existing YAML in kb/<category>/,
fill the fields honestly (especially why_models_miss_it), run
npm test && npm run eval, open a PR.
Mine your own usage. Set TOOLPLAN_LOG=toolplan.jsonl in the server
env, use the tool for a while, then npm run log-to-case toolplan.jsonl
— real ideas become eval-case skeletons; weak matches show you exactly
which keywords the KB is missing.
Run the weekly refresh. Point a Claude agent at
pipeline/REFRESH.md; it researches new tools and
writes proposals to pipeline/proposals/<date>/ with evidence. You
review PROPOSAL.md, move accepted files into kb/, run
npm test && npm run eval, commit.
Staleness check anytime: npm run stale.
npm install
npm run build
npm test
npm run smoke # end-to-end stdio call against the built serverv1: web-application scope only. See PLAN.md for roadmap (eval harness, refresh pipeline, host adapters).
Turns a raw project idea into an enriched, polished prompt using a curated knowledge base of stacks, tools, MCPs, skills, and execution directives.
Inputs: idea (required), grade (required), tags.
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
Loading tool schema changelog…
Scored across 1 tool
With only one tool, there is no possibility of confusion between tools.
Single tool named with clear verb_noun pattern, consistent within the set.
A single tool feels thin for a server that references a knowledge base, though it may be sufficient for a focused planning utility.
The tool covers the core planning action but lacks additional operations for managing or querying the underlying knowledge base.
Turn PRDs and product ideas into structured specs so coding agents build your intent, not theirs.
Turn rough ideas into structured prompts for ChatGPT, Claude, Gemini, image and video AI.
Turns vague automation requests into tool stacks, prompts, QA checks, and human boundaries.
Turn your app idea into IA, wireframes, PRD, style guides, and dev specs for coding agents.
