bonnard-data

bonnard-data / bonnard

Public

Open-source agentic schema for reliable data outputs. Query data through MCP and via our SDK. Create apps, embed data or just simply explore through your preferred agent.

bonnard.dev agentic agentic-analytics agentic-schema ai-analytics bigquery
12
1
100% credibility
Found Mar 05, 2026 at 12 stars -- GitGems finds repos before they trend. Get early access to the next one.
Sign Up Free
AI Analysis
TypeScript
AI Summary

Bonnard provides a self-hosted analytics layer that enables AI agents to query consistent, governed metrics from various data warehouses via a standardized protocol.

How It Works

1
🔍 Discover Bonnard

You hear about Bonnard, a simple way to let AI helpers pull reliable insights from your business data without relying on anyone else.

2
📥 Grab the setup kit

Download the easy starter package to run everything on your own computer or server.

3
🔗 Connect your data source

Point it to where your company numbers live, like your database, so it knows what to analyze.

4
✏️ Define your key metrics

Describe everyday business measures like total sales, customer counts, or growth rates in plain files.

5
🚀 Launch your data hub

Press go, and your private analytics engine springs to life, ready for action.

6
🖥️ Check the dashboard

Open the web page to browse your metrics, see everything working smoothly.

7
🤖 Invite AI assistants

Share a secure link so your favorite AI tools can chat with your data directly.

🎉 Get consistent insights

Now AI agents deliver trustworthy answers from your data every time, across apps and chats.

Sign up to see the full architecture

6 more

Sign Up Free

Star Growth

See how this repo grew from 12 to 12 stars Sign Up Free
Repurpose This Repo

Repurpose is a Pro feature

Generate ready-to-use prompts for X threads, LinkedIn posts, blog posts, YouTube scripts, and more -- with full repo context baked in.

Unlock Repurpose
AI-Generated Review

What is bonnard?

Bonnard is a self-hosted semantic layer built in TypeScript that delivers consistent, governed data outputs to AI agents and apps via a single schema definition. It uses Cube.js under the hood for SQL-based metrics with caching and multi-DB support (Snowflake, BigQuery, Postgres, etc.), exposed through an MCP server for agentic open source LLM queries and a simple SDK. Spin it up with Docker Compose, define models in YAML, deploy via CLI—no cloud required.

Why is it gaining traction?

It bridges data warehouses to agentic workflows seamlessly: agents like Claude or CrewAI query metrics reliably over MCP without hallucinating schemas, while the admin UI and `bon deploy` make schema updates live without restarts. As an open source agentic platform and workflow tool, it stands out for self-hosters dodging vendor lock-in, with production-ready auth, TLS via Caddy, and health checks.

Who should use this?

Data engineers defining metrics for agentic RAG or open source agentic chatbots/coding tools. Teams building internal AI apps on self-hosted infra, like querying sales data in Cursor or Claude Desktop. Avoid if you need battle-tested scale—fine for prototypes or small warehouses.

Verdict

Promising early open source agentic workflow tool at 12 stars and 1.0% credibility score—docs are solid, CLI intuitive, but low adoption signals immaturity; test thoroughly before prod. Worth a Docker spin-up for agentic data access experiments.

(178 words)

Sign up to read the full AI review Sign Up Free

Similar repos coming soon.