agent-swarm.dev
The first AI-native company operating system.
FreeOpen SourceAICode2,419 impressions#6 of its week1 comment
Comments
>log in to commentHi, I'm Taras, co-founder and CTO of Desplega Labs. We built Agent Swarm because one agent stopped being enough for our own work: research, plans, code, reviews and the follow-ups each needed different tools and a different check. So we made it a factory you run yourself. Work comes in from Slack, Linear, email or GitHub, and a lead agent turns it into tasks for coding agents like Claude Code or Codex. Your team adds the steps it cares about, like tests, a review or a risk check. Low-risk changes can then merge on their own, and risky ones wait for a person. It's MIT and runs on your own infrastructure and model keys. We run it on our own repo every day, and Capchase runs it with +60 active users. Ask me anything about the review gates, what it costs to run, or what breaks when you give agents real work.
Work goes in from Slack, GitHub, an issue tracker or the API, and a reviewed pull request comes back to your repo. A lead agent splits the goal into tasks and hands them to coding agents such as Claude Code or Codex or any other harness. Every agent runs in its own Docker container with the tools it needs.
The team keeps what it learns: shared memory, reusable skills you can edit in place, schedules that fire on their own, and a review before anything merges. Work survives a restart.
MIT, self-hosted, on your own model keys. Pick the agent per task: Claude Code, Codex, pi, opencode, Cursor, Amp, Grok, Devin, or bring your own ACP agent.
940+ GitHub stars. Capchase took it from a proof of concept to 44 active users, 26 agents and more than 10,000 tasks a week in one quarter.
Start with one command: npx skills add desplega-ai/agent-swarm
agent-swarm.dev alternatives →More Open Source tools →74 free dev tools →
- Open‑source operating system for AI work
- Lead agent delegates goals to workers like Claude Code or Codex
- Engine to make your company AI Native
- Daily evolution video showing updates
- Integration guides and extensive docs
Agent Swarm is a self-hosted, Docker-based AI software factory that turns Slack, GitHub or API requests into reviewed pull requests.
- for
- Dev teams that want on-prem AI coding agents and full control over data and models
- pricing
- open source
- license
- MIT
Key features
- Docker-isolated agents — Each coding agent runs in its own container with required tools.
- Multi-model support — Choose from Claude Code, Codex, pi, opencode, Cursor, Amp, Grok, Devin or custom agents.
- Shared memory & reusable skills — Agents keep learned context, skills and schedules across tasks.
- End-to-end workflow — Input from Slack, GitHub or API, task splitting, execution and PR review.
- Persistent state — Work survives restarts, ensuring continuity.
- MIT licensed — Open source, self-hosted with your own model keys.
Use cases
- Automatically generate and merge PRs from Slack commands
- Run scheduled code generation or refactoring tasks
- Maintain a shared knowledge base across multiple agents
- Integrate custom AI models into existing CI pipelines
agent-swarm.dev vs alternatives
 agent-swarm.dev |  LangChain |  CrewAI |  Mastra |  Devin | |
|---|---|---|---|---|---|
| Best for | Self-hosted AI coding pipelines | LLM app framework | Multi-agent AI platform | TypeScript AI agent framework | Single autonomous coding agent |
| Pricing | Open source | Free | Subscription | Free | Subscription |
| DevHunt upvotes | 17 | 0 | 0 | 0 | 0 |
| Launched | Oct 2026 | — | — | — | — |
- agent-swarm.dev vs LangChain: LangChain is a library for building LLM apps, not a ready-made self-hosted agent factory.
- agent-swarm.dev vs CrewAI: CrewAI provides a framework and platform for multi-agent systems but isn’t a self-hosted Docker solution.
- agent-swarm.dev vs Mastra: Mastra is a TypeScript framework for AI agents, focusing on code rather than a full software factory.
- agent-swarm.dev vs Devin: Devin is an autonomous AI software engineer, a single agent rather than a swarm with shared memory.
agent-swarm.dev FAQ
How do I get started?+
Run a single command: npx skills add desplega-ai/agent-swarm.
Which coding agents can I use?+
You can select Claude Code, Codex, pi, opencode, Cursor, Amp, Grok, Devin, or provide your own ACP agent.
Is the platform self-hosted?+
Yes, it is MIT-licensed and runs on your own infrastructure.
Does work survive a restart?+
Yes, the system persists tasks, memory and schedules across restarts.
How does Slack integration work?+
Agent Swarm includes a Slack manifest and can receive work items directly from Slack messages.
Summarized by DevHunt from github.com · Oct 25, 2026. Details may change; check the official site.
agent-swarm.dev in the news
Recent agent-swarm.dev news and updates from DevHunt Daily.
· DevHunt launch
Agent Swarm launches AI-native operating system for automating code tasks
Agent Swarm provides a self‑hosted system that receives work from Slack, GitHub, or an API, splits goals into tasks, and assigns them to coding agents like Claude Code or Codex. Each agent runs in its own Docker container, preserving shared memory and reusable skills across restarts, and integrates with existing repositories for automated pull‑request reviews.
