barnum-circus

barnum-circus / barnum

Public

Task queue orchestrator for AI agents

38
2
100% credibility
Found Mar 17, 2026 at 38 stars -- GitGems finds repos before they trend. Get early access to the next one.
Sign Up Free
AI Analysis
Rust
AI Summary

Barnum orchestrates AI agents through declarative workflows defined in JSON configurations, enabling complex multi-step tasks with validation, parallelism, and error handling.

How It Works

1
🔍 Discover Barnum

You hear about Barnum, a friendly ringmaster that helps tame wild AI helpers for big, tricky jobs like fixing code or checking files.

2
🚀 Call your performers

You gather your smart AI friends and get them ready to jump into action together.

3
đź“ť Sketch the show plan

You jot down a simple outline of the steps, like who does what and in what order.

4
🎪 Start the performance

With one easy go, Barnum directs your friends exactly through your plan, keeping everyone on track.

5
đź‘€ Watch it unfold

You sit back as tasks flow smoothly, friends tackling their parts perfectly without confusion.

âś… Perfect results

Your tough job is done flawlessly, with everything organized and no loose ends.

Sign up to see the full architecture

4 more

Sign Up Free

Star Growth

See how this repo grew from 38 to 38 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 barnum?

Barnum is a Rust CLI and library for orchestrating task queues as state machines tailored to AI agents, like those in Claude or GPT. You define workflows in JSON configs—linear pipelines, branching decisions, or fan-out parallelism—and it dispatches atomic tasks to a pool of long-lived agents via a file-based daemon called Troupe. This slices complex jobs into validated steps with retries, timeouts, hooks, and even local shell commands, dodging bloated contexts that make agents forgetful.

Why is it gaining traction?

Unlike loose loops like Claude's /loop or task queue Celery, Barnum enforces valid transitions, schemas, and audit trails upfront, letting you reason about workflows before runtime. Fan-out shines for parallel agent swarms (e.g., analyze 20 files at once), while hooks and command actions mix AI with deterministic glue like jq transforms—perfect for GitHub task runner setups or task queue Hubspot flows. Demos for refactoring pipelines and self-healing linters hook devs chasing reliable multi-agent automation.

Who should use this?

AI workflow builders tackling code analysis, refactors, or invariant checks across repos—think backend teams scripting GitHub task lists or markdown planning boards with agents. It's for Rust fans wanting a task queue alternative to GitHub Actions for local agent pools, especially in branching demos mimicking Barnum circus Nürnberg spectacles.

Verdict

With 38 stars and 1.0% credibility, Barnum's early—light docs, no broad tests—but the CLI (`barnum run --config workflow.jsonc`) and demos deliver instant value for agent experiments. Try for structured AI tasks; skip for production until more battle-tested.

(198 words)

Sign up to read the full AI review Sign Up Free

Similar repos coming soon.