Self-hosted · PostgreSQL · Open source

Give your agents
long-term memory.

Ingest documents. Extract knowledge graphs. Store temporal entities and relations. Query with bounded traversal. Your agents remember what they learned across sessions.

PDF MD HTML API S3 CSV Log Git User entity Auth module Token service Config store Route handler DB adapter OGM core uses authenticates generates reads
Extraction Pipeline

From documents to knowledge.

A durable pipeline that ingests, parses, chunks, extracts, and reconciles. Automatic retries on failure.

01

Ingest

Upload PDF, Markdown, HTML, or plain text. Size, type, and duplicate validation.

02

Parse & Chunk

Extract text content. Split into semantic chunks preserving context boundaries.

03

Extract

LLM-powered entity, relation, and evidence extraction from each chunk.

04

Reconcile

Deduplicate entities. Merge temporal versions. Build the knowledge graph.

Harness Integration

Connect your agents.

Works with Claude Code, OpenCode, and Hermes through MCP protocol. One bridge, multiple harnesses.

Claude Code Anthropic OpenCode SVC Labs Hermes Agent Framework MCP MCP MCP ogm-agent-bridge MCP Bridge 20 tools · 3 profiles pip install ogm-agent-bridge REST open-graph-memory OGM Core Knowledge Graph Agent Memory FastAPI + PostgreSQL opengraphmemory-skill workflow definition defines
CC

Claude Code

Anthropic's coding agent. Configure via .claude/mcp.json.

OC

OpenCode

SVC Labs' coding agent. Configure via opencode.json.

HM

Hermes

Agent framework. Configure via hermes.config.json.

Ecosystem

Three projects, one stack.

The skill tells agents what to do. The bridge connects them. The core stores everything.

Workflow Definition

opengraphmemory-skill

Tells AI agents how and when to use Agent Memory. Defines the recall, reorient, and persist protocol. Not a runtime — a behavioral contract.

View repo
MCP Transport

ogm-agent-bridge

Stateless MCP bridge connecting Claude Code, OpenCode, and Hermes to OpenGraphMemory REST APIs. 20 tools, 3 permission profiles. Published on PyPI.

View repo
Core Platform

open-graph-memory

Knowledge graph extraction, temporal storage, and Agent Memory API. FastAPI + PostgreSQL. The backend that powers everything.

View repo
Agent Memory

Episodes, not chat logs.

Structured memory with temporal supersession. Agents record what they learned, not what they said.

Episodes

Immutable observations of what happened. Errors encountered, fixes applied, patterns discovered.

Attempts

Hypotheses tested within an episode. What was tried, what happened, what was the result.

Outcomes

Verified results with verifiers. A test passes, a build succeeds, a deploy completes.

Patterns

Distilled knowledge from verified outcomes. Durable conclusions that supersede earlier understanding.

Time Travel

Query the knowledge graph at any point in time. See how entities and relations evolved.

Permission Profiles

Read-only by default. Personal-safe for writes. Memory-curator for curation governance.

Stack

Built with.

Python 3.12+
FastAPI
PostgreSQL 16+
Redis + ARQ
React + Vite
Tailwind CSS
S3 Compatible
Docker Compose
Alembic
NetworkX
Get Started

Up in minutes.

terminal
# Clone and start git clone https://github.com/ardiannurcahya/open-graph-memory.git cd open-graph-memory cp .env.example .env docker compose up -d # API at http://localhost:8000 # Web UI at http://localhost:5173 # Connect agents via MCP pip install ogm-agent-bridge uvx ogm-agent-bridge # runs as MCP stdio server