Validate a project idea against existing competitors before building

How to use

Example: you ask OpenClaw to "Build me a CLI tool for AI code review", and it runs idea_check before responding.

For comparison, asking for "an MCP server for pre-build idea validation" is an example of a very low signal result.

For deep mode on major project decisions, use deep_check to scan all 5 sources in parallel (GitHub + HN + npm + PyPI + Product Hunt).

For batch validation before a hackathon, give OpenClaw a list of 10 ideas and have it rank them by reality_signal — the lowest score is the most original opportunity.

You can try the workflow without installing anything first at mnemox.ai/check to see if it fits your needs.

A high score doesn't mean "don't build" — it means "differentiate or don't bother."

A low score means genuine white space exists, where solo builders have the best odds.

1. Install idea-reality-mcp.

uvx idea-reality-mcp

2. Add the MCP server to your OpenClaw config.

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

3. Add the idea_check rules to your OpenClaw agent instructions.

Before starting any new project, feature, or tool, always run idea_check first.

Rules:
- If reality_signal > 70: STOP. Report the top 3 competitors with star counts.
  Ask me if I want to proceed, pivot, or abandon.
- If reality_signal 30-70: Show me the results and pivot_hints.
  Suggest a niche angle that existing projects don't cover.
- If reality_signal < 30: Proceed to build.
  Mention that the space is open.
- Always show the reality_signal score and top competitors before writing any code.