Let your agent kill the idea.

Let your agent set it up

Paste this into Claude Code, Codex, Cursor or any coding agent. It reads the Markdown docs and configures itself.

Add the Kill My Idea MCP server to my setup. It is a free, public remote MCP server (Streamable HTTP) at https://killmyidea.biz/api/mcp with no API key and no auth. Read https://killmyidea.biz/mcp.md for the exact config for the client you are running in, install it, then list its tools to confirm it works.

Install

One URL. Pick your client.

claude mcp add --transport http killmyidea https://killmyidea.biz/api/mcp

{
  "mcpServers": {
    "killmyidea": { "url": "https://killmyidea.biz/api/mcp" }
  }
}

{
  "servers": {
    "killmyidea": { "type": "http", "url": "https://killmyidea.biz/api/mcp" }
  }
}

codex mcp add killmyidea --url https://killmyidea.biz/api/mcp

[mcp_servers.killmyidea]
url = "https://killmyidea.biz/api/mcp"

{
  "mcpServers": {
    "killmyidea": { "serverUrl": "https://killmyidea.biz/api/mcp" }
  }
}

gemini mcp add --transport http killmyidea https://killmyidea.biz/api/mcp

{
  "mcpServers": {
    "killmyidea": { "command": "npx", "args": ["-y", "mcp-remote", "https://killmyidea.biz/api/mcp"] }
  }
}

Tools

What your agent shows in the terminal (real output)

KILL IT   ████████▎░░░░░░░░░░░  41/100
Social · goal: Make money
──────────────────────────────────────────────
Real problem    ███▋░░░░░░░░░░░░░░░░   18  ×2
Clear customer  ███████░░░░░░░░░░░░░   35
Demand          ███████▍░░░░░░░░░░░░   37
Money           ██▋░░░░░░░░░░░░░░░░░   13  ×2
Reach           ███████████░░░░░░░░░   55
Different       █████████████████▍░░   87
Buildable       █████████████████▍░░   87
Shareable       █████████▍░░░░░░░░░░   47
──────────────────────────────────────────────
+ Different 87   It doesn't look like everything else.
- Money 13       Getting users may be easier than getting customers.

“A social network where you can only post once a year, on your birthday.”

The same result as structuredContent (trimmed)

{
  "verdict": "FIX",
  "verdict_label": "FIX IT",
  "score": 58,
  "needs_detail": false,
  "goal": "money",
  "category": "SaaS",
  "best_signal": { "dimension": "buildable", "label": "Buildable", "score": 75,
                   "note": "You could ship a first version this week." },
  "biggest_risk": { "dimension": "money", "label": "Money", "score": 25,
                    "note": "Getting users may be easier than getting customers." },
  "dimensions": [
    { "key": "problem", "label": "Real problem", "score": 75, "weight": 2,
      "question": "Does it fix something people actually feel?" },
    …
  ],
  "thresholds": { "kill": "<50", "fix": "50-64", "ship": "65+" },
  "scoring_version": 2,
  "latency_ms": 912
}

When needs_detail is true, the idea was too vague to judge: verdict and score are null. A good agent asks the user for more detail instead of guessing.

Build it into your own agent

Raw HTTP

curl -s https://killmyidea.biz/api/mcp \
  -H 'Content-Type: application/json' \
  -H 'Accept: application/json, text/event-stream' \
  -d '{"jsonrpc":"2.0","id":1,"method":"tools/call","params":{
        "name":"judge_idea",
        "arguments":{"idea":"A Slack bot that turns standups into a weekly changelog for clients.","goal":"money"}}}'

TypeScript MCP SDK

import { Client } from '@modelcontextprotocol/sdk/client/index.js'
import { StreamableHTTPClientTransport } from '@modelcontextprotocol/sdk/client/streamableHttp.js'

const client = new Client({ name: 'my-agent', version: '1.0.0' })
await client.connect(new StreamableHTTPClientTransport(new URL('https://killmyidea.biz/api/mcp')))

const { structuredContent } = await client.callTool({
  name: 'judge_idea',
  arguments: { idea: 'A Slack bot that turns standups into a weekly changelog for clients.' },
})

Claude API (MCP connector)

curl https://api.anthropic.com/v1/messages \
  -H "x-api-key: $ANTHROPIC_API_KEY" \
  -H 'anthropic-version: 2023-06-01' \
  -H 'anthropic-beta: mcp-client-2025-11-20,server-side-fallback-2026-07-01' \
  -H 'content-type: application/json' \
  -d '{
    "model": "claude-opus-5-5",
    "max_tokens": 16000,
    "fallbacks": "default",
    "mcp_servers": [{ "type": "url", "url": "https://killmyidea.biz/api/mcp", "name": "killmyidea" }],
    "tools": [{ "type": "mcp_toolset", "mcp_server_name": "killmyidea" }],
    "messages": [{ "role": "user", "content": "Judge my idea: a Slack bot that turns standups into a weekly changelog for clients." }]
  }'

OpenAI Responses API

"tools": [{
  "type": "mcp",
  "server_label": "killmyidea",
  "server_url": "https://killmyidea.biz/api/mcp",
  "require_approval": "never"
}]

How Jev decides

No chatbot opinion. Jev by TypeSafe AI answers ten fixed questions in parallel with calibrated probabilities: eight indie-hacker questions scored 0-4, the category, and whether the idea is understandable at all.

• Each answer × 25 → 0-100. The score is the weighted average. Real problem and Money count twice.
• KILL IT below 50 · FIX IT 50-64 · SHIP IT 65+.
• The same idea gets the same rubric every time. The code is open source.

FAQ

Is it really free?

Yes. No API key, no account, no OAuth. Please don't hammer it: it is a side project.

Which clients work?

Anything that supports remote MCP servers over Streamable HTTP. Older clients can bridge with npx mcp-remote.

Are my ideas stored?

Archived anonymously for product analytics, with no IP address and no identifiers. Pass do_not_archive: true and the text is not stored. Calls are counted (tool name, client name, user agent), never their arguments.

Can I trust the verdict?

It is a fast, consistent first filter, not market research. Use the biggest risk to decide what to test next.