connect_sources
Connect live sources — data stays where it lives
Open source · Ontology AI · Reasoning As Code
AnythingGraph — the reasoning layer between agents and your databases.
Ontology AI for enterprise data — write ownership, relationships, and access rules once. Agents ask in plain English; AnythingGraph traverses live systems and returns answers you can audit. No data copy. No prompt stuffing.
Claude
Claude
How it works
Four steps from live data to governed agent answers — no warehouse dump, no custom glue per model.
explore_schemas
Explore schemas across your CRM, database, documents, ERP, and more
ontology_as_code
Let your AI agent draft the ontology playbook — you approve
use_in_ai_agents
Agents ask in plain English — get scoped, governed answers
Why AI developers love AnythingGraph
Anything Graph is the MCP layer between your AI agents and siloed enterprise data — secure access, context, and orchestration without moving a single record.
Use your favorite AI agent
Claude, OpenAI, Cursor, and other AI tools connect in minutes — no custom setup for each model.
Connect your data without moving it
Salesforce, Postgres, MySQL, MongoDB, REST APIs, CSV files, and more. Answers come from your live systems — records never leave.
An Ontology layer that understands your business
Anything CLI sits between your agent and siloed systems — turning everyday questions into safe, accurate answers.
Lightweight, secure, scalable
Fast, secure, and built to scale — with access rules enforced on every answer.
Your rules, written once, with governance.
Describe your business once — who sees what — and Anything CLI maps it to your real systems.
Complete open source
Completely open source. One command to get started — fork it, extend it, run it your way.
Stop dump everything into the prompt
With governed queries over live systems — less spend, less rework, and answers you can explain.
Lower token spend
Query only what the playbook needs — not millions of warehouse rows or document chunks in every prompt.
Not months of integration
Install, map a playbook to your sources, and connect MCP. Skip the data-lake and ETL project before your first agent answer.
Data copy cost
Records stay in Postgres, Salesforce, CSV, and APIs where they already live — no migration tax or sync pipeline to maintain.
Fewer bad answers
Early teams report cutting agent hallucinations when answers route through playbooks, bindings, and proof instead of guesswork.
Rules, many agents
Write access and vocabulary once in a playbook — reuse the same governed layer for Claude, Cursor, OpenAI, and other MCP clients.
On every reply
Source records and relationship paths come back with the answer — less audit friction and less time re-checking the agent by hand.
One-step setup
Install Anything CLI, set up your access rules, and connect your AI tool — done in minutes, not hours. No copying data, no heavy integration project.
Three steps and your agent speaks your business language — with the right people seeing the right data.
- Install
- Set up your access rules
- Connect your AI tool
Step-by-step: explore Postgres + MongoDB and build a cross-source CRM playbook via MCP →
Connect where your data already lives
Anything CLI uses adapters to query systems in place — no copy step. These adapters ship today with the same playbook and binding model.
PostgreSQL
sql
CSV files
csv
Salesforce
soql
MySQL
mysql
SQL Server
mssql
MongoDB
mongodb
REST / OpenAPI
rest
Ready to connect your agent to live data?
Anything CLI is open source on GitHub. Install, connect your AI tool, and ask your first question — in minutes, not hours.