Graphonomous Documentation

Part of the [&] Protocol stack · Ecosystem overview · Three-protocol stack · Stack status

Welcome to the documentation hub for Graphonomous.

Graphonomous is a continual-learning memory engine for AI agents, implemented in Elixir/OTP and exposed via MCP tools/resources. It combines graph-based memory, outcome-driven confidence updates, goal orchestration, topology-aware routing, and consolidation cycles.

Start Here

If you're new to the project, begin with:

1. Quickstart — run Graphonomous locally and verify the memory loop.

2. Architecture — understand runtime components and data flow.

3. Runtime Walkthrough — follow retrieve → act → store → learn in practice.

4. MCP Tools — reference all tools and resources.

5. Operations — runbook for maintenance and troubleshooting.

Documentation

Homepages

• [&] Ampersand Box

• Graphonomous

• BendScript

• WebHost.Systems

• Agentelic

• AgenTroMatic

• Delegatic

• Deliberatic

• FleetPrompt

• GeoFleetic

• OpenSentience

• SpecPrompt

• TickTickClock

Root Docs

• [&] Protocol Docs

• Graphonomous Docs

• BendScript Docs

• WebHost.Systems Docs

• Agentelic Docs

• AgenTroMatic Docs

• Delegatic Docs

• Deliberatic Docs

• FleetPrompt Docs

• GeoFleetic Docs

• OpenSentience Docs

• SpecPrompt Docs

• TickTickClock Docs

Graphonomous Docs

• quickstart

• architecture

• runtime-walkthrough

• mcp-tools

• operations

• faq

• README

Integration & Build Guides

Setup & Distribution

• BOOTSTRAP

• ZED

• NPM_PUBLISH

Agent Skills Pack

Skills

• SKILLS

• bootstrap

• retrieve

• store

• learn

• deliberate

• consolidate

• goals

• belief

• forgetting

• epistemic-frontier

• trace-evidence-path

• attention

• review

• inspect

• graph-health

• workflows

• sync

• watch

Documentation index

Linked index of every page in this set (renders on GitHub and in the docs atlas; the toctrees above drive the Sphinx build):

Guides

• Architecture

• Quickstart

• Runtime Walkthrough

• MCP Tools Reference

• Operations & Maintenance

• FAQ

• Technical Documentation

• Arithmetic Compliance Review (box-and-box)

• Agent Skills

Setup & distribution

• Local Bootstrap & Verification Guide

• Zed MCP Setup Guide

• NPM Package Specification

• npm Publishing + Maintenance Runbook

Specification

• κ Integration Spec — Product-Level Specification

• κ × Product Crosswalk

• κ-Aware Graph Intelligence — Theoretical Foundation

UX

• User Stories

Build prompts

• Benchmark Expansion Prompt

• Continual Learning v0.3 — Implementation Prompt

• Continual Learning v0.3 — P1 (Week 3-4)

• Continual Learning v0.3 — P2 (Week 5+)

• Continual Learning v0.3 — P3 (Week 6+)

• κ-Topology QA & SHR Optimization — Session Prompt

Recommended Reading Path

• Operators/Maintainers: quickstart → operations → NPM_PUBLISH

• Agent Integrators: mcp-tools → runtime-walkthrough → skills/SKILLS

Core Operating Loop

For non-trivial work, use this cycle consistently:

1. Retrieve context.

2. Reason and act.

3. Store durable knowledge.

4. Learn from outcomes.

5. Consolidate periodically.

This loop keeps Graphonomous memory accurate, adaptive, and useful across sessions.