Back to Blog
QuickstartApril 7, 20265 min read

MCP Integration in 5 Minutes

NG

Nick Gallick

Founder

What is MCP?

The Model Context Protocol (MCP) is an open standard for connecting AI models to external tools and data sources. Hipp0 provides a full MCP server that exposes 12+ tools for decision memory operations. Any MCP-compatible agent (Claude, GPT, custom agents) can connect to Hipp0 and immediately start recording, querying, and reasoning about decisions.

Quick Setup

Add the following to your MCP configuration file. If you are using Claude Desktop, this goes in claude_desktop_config.json. For other clients, adapt to your MCP client's config format. 

{
  "mcpServers": {
    "hipp0": {
      "command": "npx",
      "args": ["-y", "@hipp0/mcp"],
      "env": {
        "HIPP0_API_KEY": "your-api-key",
        "HIPP0_PROJECT_ID": "your-project-id"
      }
    }
  }
}

Available Tools (12+)

Once connected, your agent has access to the full Hipp0 tool suite. Here are the most commonly used tools:

  • record_decision — Record a new decision with context, rationale, and tags
  • query_decisions — Retrieve relevant decisions using 5-signal scoring
  • get_context — Get compiled H0C context for a specific task or scope
  • detect_contradictions — Check if a proposed decision contradicts existing ones
  • list_decisions — List decisions with filtering and pagination
  • update_decision — Update an existing decision's status or metadata
  • supersede_decision — Mark a decision as superseded by a new one
  • get_decision_graph — Retrieve the relationship graph for a decision
  • start_session — Start a Super Brain session for multi-step workflows
  • checkpoint_session — Create a session checkpoint with compressed context

SDK Examples

If you prefer programmatic access, Hipp0 provides TypeScript and Python SDKs.

// TypeScript
import { Hipp0Client } from "@hipp0/sdk";

const hipp0 = new Hipp0Client({
  apiKey: process.env.HIPP0_API_KEY,
  projectId: "my-project",
});

const decisions = await hipp0.query({
  text: "How should we handle authentication?",
  persona: "backend-agent",
  limit: 5,
});
# Python
from hipp0 import Hipp0Client

client = Hipp0Client(
    api_key=os.environ["HIPP0_API_KEY"],
    project_id="my-project",
)

decisions = client.query(
    text="How should we handle authentication?",
    persona="backend-agent",
    limit=5,
)
All posts