Add Help an Agent MCP server (human judgment for AI agents)

[jrcruciani]

What does this server do?

Help an Agent lets AI agents pause and consult real humans when they face decisions requiring emotional, social, ethical, or cultural judgment — situations where technical correctness isn't enough.

Tools provided:

• ask_humans — Submit a question with context, wait for human consensus (polls with backoff)
• check_pulse — Check status of a previously submitted consultation

How it works:

1. Agent calls ask_humans with a question and category
2. Real humans respond on helpanagent.site with direction (yes/no/depends) and certainty (1–5)
3. Weighted consensus engine aggregates responses (reputation-weighted, outlier detection, 25% individual cap)
4. Agent gets back: consensus direction, confidence score, response count

Stack: Cloudflare Workers + D1 + KV (fully serverless)
npm: helpanagent-mcp
Website: https://helpanagent.site
Docs: https://helpanagent.site/docs
GitHub: https://github.com/jrcruciani/helpanagent

Installation

{
  "mcpServers": {
    "helpanagent": {
      "command": "npx",
      "args": ["-y", "helpanagent-mcp"],
      "env": {
        "HELPANAGENT_API_KEY": "your_api_key"
      }
    }
  }
}