English | 繁體中文

idea-reality-mcp

We search. They guess.

The only idea validator that searches real data. 5 sources. Quantified signal. Zero hallucination.

The problem

Every developer has wasted days building something that already exists with 5,000 stars on GitHub.

You ask ChatGPT: "Is there already a tool that does X?"

ChatGPT says: "That's a great idea! There are some similar tools, but you can definitely build something better!"

That's not validation. That's cheerleading.

What we do instead

You: "AI code review tool"

idea-reality-mcp:
├── reality_signal: 90/100
├── GitHub repos: 847
├── Top competitor: reviewdog (9,094 ⭐)
├── npm packages: 56
├── HN discussions: 254
└── Verdict: HIGH — consider pivoting to a niche

One gives you encouragement. The other gives you facts.

Which one do you trust your next 3 months on?

Try it now (30 seconds)

uvx idea-reality-mcp

Or try it in your browser — no install required.

Why not just ask ChatGPT?

TL;DR — We search 5 real databases. They generate opinions.

New: AI-powered search intelligence

Claude Haiku 4.5 now generates optimal search queries from your idea description — in any language — with automatic fallback to our battle-tested dictionary pipeline.

The LLM understands your idea. The dictionary is your safety net. You always get results.

Install (5 minutes)

Claude Desktop

Paste into ~/Library/Application Support/Claude/claude_desktop_config.json (macOS) or %APPDATA%\Claude\claude_desktop_config.json (Windows):

{
  "mcpServers": {
    "idea-reality": {
      "command": "uvx",
      "args": ["idea-reality-mcp"]
    }
  }
}

Cursor

Paste into .cursor/mcp.json in your project root:

{
  "mcpServers": {
    "idea-reality": {
      "command": "uvx",
      "args": ["idea-reality-mcp"]
    }
  }
}

Claude Code (CLI)

claude mcp add idea-reality -- uvx idea-reality-mcp

Optional: Environment variables

export GITHUB_TOKEN=ghp_...        # Higher GitHub API rate limits
export PRODUCTHUNT_TOKEN=your_...  # Enable Product Hunt (deep mode)

Optional: Agent auto-trigger

The MCP tool description already tells your agent what idea_check does. To make it run proactively (before every new project), add a one-line hint:

When starting a new project, use the idea_check MCP tool to check if similar projects already exist.

Add to your CLAUDE.md, .cursorrules, .windsurfrules, or .github/copilot-instructions.md. See templates/ for all platforms.

Usage

"I have a side project idea — should I build it?"

Tell your AI agent:

Before I start building, check if this already exists:
a CLI tool that converts Figma designs to React components

The agent calls idea_check and returns: reality_signal, top competitors, and pivot suggestions.

"Find competitors and alternatives"

idea_check("open source feature flag service", depth="deep")

Deep mode scans all 5 sources in parallel — GitHub repos, HN discussions, npm packages, PyPI packages, and Product Hunt — and returns ranked results.

"Build-or-buy sanity check before a sprint"

We're about to spend 2 weeks building an internal error tracking tool.
Run a reality check first.

If the signal comes back at 85+ with mature open-source alternatives, you just saved your team 2 weeks.

Tool schema

idea_check

Output: reality_signal (0-100), duplicate_likelihood, evidence[], top_similars[], pivot_hints[], meta{}

{
  "reality_signal": 72,
  "duplicate_likelihood": "high",
  "evidence": [
    {"source": "github", "type": "repo_count", "query": "...", "count": 342},
    {"source": "github", "type": "max_stars", "query": "...", "count": 15000},
    {"source": "hackernews", "type": "mention_count", "query": "...", "count": 18},
    {"source": "npm", "type": "package_count", "query": "...", "count": 56},
    {"source": "pypi", "type": "package_count", "query": "...", "count": 23},
    {"source": "producthunt", "type": "product_count", "query": "...", "count": 8}
  ],
  "top_similars": [
    {"name": "user/repo", "url": "https://github.com/...", "stars": 15000, "description": "..."}
  ],
  "pivot_hints": [
    "High competition. Consider a niche differentiator...",
    "The leading project may have gaps in...",
    "Consider building an integration or plugin..."
  ],
  "meta": {
    "sources_used": ["github", "hackernews", "npm", "pypi", "producthunt"],
    "keyword_source": "llm",
    "depth": "deep",
    "version": "0.4.0"
  }
}

Scoring weights

If Product Hunt is unavailable (no token), its weight is redistributed automatically.

CI: Auto-check on Pull Requests

Use idea-check-action to validate new feature proposals:

name: Idea Reality Check
on:
  issues:
    types: [opened]

jobs:
  check:
    if: contains(github.event.issue.labels.*.name, 'proposal')
    runs-on: ubuntu-latest
    steps:
      - uses: mnemox-ai/idea-check-action@v1
        with:
          idea: ${{ github.event.issue.title }}
          github-token: ${{ secrets.GITHUB_TOKEN }}

Roadmap

• [x] v0.1 — GitHub + HN search, basic scoring
• [x] v0.2 — Deep mode (npm, PyPI, Product Hunt), improved keyword extraction
• [x] v0.3 — 3-stage keyword pipeline, 150+ Chinese term mappings, synonym expansion, LLM-powered search (Render API)
• [x] v0.4 — Email gate, Score History, Agent Templates, GitHub Action
• [ ] v0.5 — Temporal signals (trend detection and timing analysis)
• [ ] v1.0 — Idea Memory Dataset (opt-in anonymous logging)

Found a blind spot?

If the tool missed obvious competitors or returned irrelevant results:

1. Open an issue with your idea text and the output
2. We'll improve the keyword extraction for your domain

License

MIT — see LICENSE

Contact

Built by Mnemox AI · dev@mnemox.ai