lostbean

lostbean / jido_composer

Public

Composable agent flows the Jido ecosystem

22
3
100% credibility
Found Mar 19, 2026 at 22 stars -- GitGems finds repos before they trend. Get early access to the next one.
Sign Up Free
AI Analysis
Elixir
AI Summary

Jido Composer is an Elixir library for creating composable agent workflows that blend deterministic state machines, LLM-driven orchestrators, human-in-the-loop approvals, and durable persistence.

How It Works

1
🔍 Discover smart workflows

You hear about a tool that lets you build reliable pipelines mixing step-by-step tasks, smart AI decisions, and human checks for things like code reviews.

2
📦 Add to your project

Simply add it to your Elixir app's dependencies and you're ready to create powerful flows.

3
🛠️ Design your pipeline

Describe your process like parallel checks for linting, security scans, and tests, with a human approval step before merging.

4
🤖 Mix AI and humans seamlessly

Nest smart AI thinkers inside fixed steps, add pauses for your review, and it all flows together naturally.

5
▶️ Run on a pull request

Feed in a repo and PR details, watch it run checks, pause for your okay, then resume to finish.

✅ Automated reviews live

Your code review pipeline now handles PRs reliably with AI smarts, your oversight, and safe merges every time.

Sign up to see the full architecture

4 more

Sign Up Free

Star Growth

See how this repo grew from 22 to 22 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 jido_composer?

jido_composer is a composable agent architecture in Elixir for the Jido ecosystem, letting you build agent flows that mix deterministic FSM workflows with LLM-driven orchestrators. Nodes like actions, agents, fan-outs, and human gates nest arbitrarily via a uniform context-passing interface, enabling pipelines from ETL to adaptive reasoning. It handles human-in-the-loop approvals and durable checkpoint/resume out of the box, perfect for long-running composable agents.

Why is it gaining traction?

Its deep composability stands out—nest workflows inside orchestrators or vice versa without glue code—while baked-in persistence and HITL avoid common pitfalls in agentic AI setups. Developers hook on the declarative DSL for quick pipelines (e.g., code review with parallel linting/security scans) and sync run/query APIs that just work. In the Elixir ecosystem, it bridges structured flows and AI agents seamlessly.

Who should use this?

Elixir backend devs building agentic platforms like automated reporting pipelines or multi-step customer support bots needing human oversight. Ideal for teams mixing rule-based ETL with LLM steps, such as devops automating PR reviews or finance apps with approval gates.

Verdict

Promising early composable agentic system with excellent docs, livebooks, and tests, but low maturity (22 stars, 1.0% credibility score) means watch for production stability. Try the examples if you're in Elixir agents—low risk, high upside.

(178 words)

Sign up to read the full AI review Sign Up Free

Similar repos coming soon.