FORGE FORGE

Audit and upgrade multi-agent systems across your coding projects

Quick Start

This clones a repo and starts an AI agent that reads instructions from it. We run automated security scans on every release — see details.
# Clone next to your projects (or where you'll start new ones)
git clone --branch v0.4 https://github.com/skondrashov/agent-forge.git

# Start it up
cd agent-forge
claude go

Pinned to a tagged release. Review the repo's CLAUDE.md to see what the agent will do.

What It Is

Agent Forge is a template agent system manager. You register your projects, and the forge scans them against a pattern library, identifies gaps, and propagates best practices — one project at a time.

How It Works

  1. Audit — scans every registered project's agent system
  2. Review — compares against the pattern library, ranks by priority
  3. Propagate — applies upgrades to one project at a time
  4. Update — refreshes patterns and the registry
  5. Repeat

Forge Roles

Forgemaster

forgemaster.md

Coordinates the audit-and-propagate cycle. Decides what to scan, what to upgrade, and in what order.

Assayer

assayer.md

Scans projects against the pattern library. Produces gap analyses with maturity levels and concrete upgrade plans.

Smith

smith.md

Applies upgrades to target projects. Adapts templates to each project's domain, conventions, and tone.

Keeper

keeper.md

Maintains the pattern library. Challenges the assayer's findings to keep the forge honest about itself.

Pattern Library

The forge scans against patterns from Rhizome, the agent orchestration pattern catalog.

Steward One agent, one memory file. The right default for almost every project. Startup Protocol Standalone PROTOCOL.md with timestamps, forum voting, and ref doc routing Shutdown Reflection Agents evaluate their context at session end; feedback fixes the docs
Reference Doc Splitting Slim AGENTS.md + role-specific ref docs to reduce context bloat
Keeper Feedback Loop A doc-maintenance agent that processes reflection feedback and keeps docs current

Security

When you run Agent Forge, your AI agent reads the role files and patterns in this repo as instructions. That's a supply chain risk — a compromised prompt file could instruct your agent to exfiltrate data or modify code in harmful ways. We take that seriously.

What we have in place:

Read the full policy: SECURITY.md