stoneforge-ai

stoneforge-ai / stoneforge

Public

A web dashboard and runtime for orchestrating AI coding agents

stoneforge.ai agent-framework agents ai ai-agents event-sourcing
23
3
100% credibility
Found Feb 25, 2026 at 22 stars -- GitGems finds repos before they trend. Get early access to the next one.
Sign Up Free
AI Analysis
TypeScript
AI Summary

Stoneforge is a web dashboard that coordinates multiple AI coding agents to collaborate on software development tasks without conflicts.

How It Works

1
🔍 Discover Stoneforge

You find Stoneforge while looking for a way to make AI helpers work together on your coding projects.

2
📥 Add to your project

You add Stoneforge to your coding project with a simple setup command.

3
🚀 Launch the dashboard

You start the web dashboard which opens in your browser to manage everything.

4
🤖 Build your AI team

You create a planner AI to organize work and worker AIs to do the coding tasks.

5
💬 Give your goal

You chat with the planner AI and describe what you want to build, like a new feature.

6
👀 Watch them collaborate

You see all AIs working together in real-time on the dashboard, no more conflicts.

✅ Get your code

Your project gets updated with clean, tested code ready to use.

Sign up to see the full architecture

5 more

Sign Up Free

Star Growth

See how this repo grew from 22 to 23 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 stoneforge?

Stoneforge delivers a TypeScript-powered web dashboard and runtime for running teams of AI coding agents in parallel. Install the CLI with npm, hit sf init and sf serve in your repo, then manage everything from localhost:3457—kanban tasks, real-time agent terminals, metrics charts, and docs search. It handles planning, isolated git worktrees, auto-merges, and handoffs so agents collaborate without conflicts, turning chaotic terminals into a unified web dashboard framework.

Why is it gaining traction?

It crushes single-agent limits and Claude's experimental teams with persistent event-sourced storage, daemon-driven task dispatch, and merge stewards that run tests before squashing PRs. Devs dig the React-fresh UI for inbox triage, playbook workflows, and multi-provider support (Claude Code, OpenAI)—no lock-in, just scalable parallelism. As a web dashboard builder akin to github web ide flows, it makes agent orchestration feel like a github web interface on steroids.

Who should use this?

Engineers scaling 3+ AI agents for feature builds, like backend teams prototyping APIs or fullstack devs handling parallel refactors. Perfect if you're juggling Claude instances across plans and need visibility beyond tmux—think solo founders or small AI-first teams ditching manual git dances for automated coordination.

Verdict

Worth npm installing for multi-agent pioneers despite 22 stars and 1.0% credibility—quickstart shines, docs are LLM-optimized, but expect token burn and sharp edges in this experimental phase. Fork it if you're stoneforged into AI coding; otherwise, wait for polish.

(198 words)

Sign up to read the full AI review Sign Up Free

Similar repos coming soon.