This project provides an automated "reality check" for AI coding agents, preventing redundant development by scanning multiple real-world databases before an agent writes code. It targets developers and AI agents seeking to validate ideas against existing solutions, offering a quantifiable reality signal and evidence from sources like GitHub, Hacker News, npm, PyPI, and Product Hunt, thereby saving significant development time and resources.

How It Works

The tool functions as an MCP (Meta-Command Protocol) client that integrates directly into AI coding agent workflows. It automatically queries five distinct data sources—GitHub repositories, Hacker News discussions, npm packages, PyPI packages, and Product Hunt launches—to assess the novelty and competitive landscape of a given idea. The core advantage lies in its automation; unlike manual searches, it triggers proactively within the agent, providing a consolidated, scored output (0-100 reality signal) with verifiable evidence, rather than raw search results.

Quick Start & Requirements

• Primary install: uvx idea-reality-mcp
• Prerequisites: uvx tool. Optional environment variables GITHUB_TOKEN and PRODUCTHUNT_TOKEN enhance API rate limits and enable full Product Hunt scanning.
• Integration: Configuration examples are provided for Claude Desktop, Claude Code, Cursor, and Smithery. A GitHub Action (mnemox-ai/idea-check-action@v1) is available for CI integration.
• Demo: A browser-based demo is available at mnemox.ai/check.

Highlighted Details

• Scans five comprehensive data sources: GitHub, Hacker News, npm, PyPI, and Product Hunt.
• Generates a single, objective "reality signal" score (0-100) indicating market saturation.
• Offers "quick" (GitHub + HN) and "deep" (all five sources) scanning modes.
• Features AI-powered search intelligence for generating optimal queries from idea descriptions.

Maintenance & Community

The project is built by Mnemox AI, with contact available at dev@mnemox.ai. Future roadmap items include "Temporal signals" (v0.5) and an "Idea Memory Dataset" (v1.0).

Licensing & Compatibility

The project is released under the MIT License, permitting commercial use and integration into closed-source projects without significant restrictions.

Limitations & Caveats

Full functionality of the "deep" scan mode, particularly concerning Product Hunt data, may depend on providing an API token. The scoring mechanism is intentionally simple and rule-based, not employing machine learning, which might limit nuanced interpretation in highly complex scenarios. Several advanced features are still on the roadmap.