workflowstacks
automation

AgentLoom

Simple, flexible workflow orchestration for multi-agent AI apps, with YAML configuration, runtime safety, observability, and resume support.
169 stars17 forksPythonUpdated 10/22/2026100% free · open source
What it does

AgentLoom lets you stitch together multiple AI agents into a reliable, resumable workflow using a simple YAML file.

When to use it
  • •You need a sequence where one LLM agent’s output becomes the input for another (e.g., research → write → edit).
  • •You want built‑in safety checks and automatic retries if an LLM call fails.
  • •You need to pause a long‑running multi‑agent job, restart it later, and keep all intermediate results.
  • •You want a single source‑of‑truth config you can version‑control for complex AI pipelines.
Ready-to-paste prompt
```bash
python -m agentloom run workflow.yaml
```
(This executes the YAML‑defined multi‑agent pipeline shown above, producing a finished blog post and persisting logs for later inspection.)
Heads up: AgentLoom aborts if the `OPENAI_API_KEY` env variable (or the key for whichever LLM you use) isn’t set before running; also, it requires Python 3.9+ and the `pyyaml` package, which is installed automatically but will fail on older Python versions.
Saves to your device
Use with Claude
New

Skip the builder — one click puts this in Claude, Cursor, Antigravity and more.

✅ Light setup

Installs with a command or two; your AI agent can do it for you.

Try it instantly — no install
Claude Code
mkdir -p ~/.claude/skills/agentloom && curl -fsSL https://workflowstacks.com/api/skills/agentloom/claude-skill -o ~/.claude/skills/agentloom/SKILL.md
Open in another AI app

Opens the app with this repo with the prompt ready to go — no copy-paste needed.

Connect the whole catalog (MCP)
claude mcp add --transport http workflowstacks https://workflowstacks.com/api/mcp

Adds a WorkflowStacks connector to Claude Code: search and load any skill here by chatting.

How AgentLoom works
Codeflow
Free to inspect

AgentLoom is a very large Python project (~110k lines across 386 code files, plus 245 test files). Setup is light: installs like a normal app. Reading the code is optional. Last commit this month, no license file, has a test suite.

Size
Very large codebase
~110k lines · 386 code files · days to read — use, don't read
Setup
One-command install
Installs like a normal app. Reading the code is optional.
Runs on
Python
No API keys detected
Python 97%TypeScript 1%Tree-sitter Query 1%
Where to start reading
  1. 1
    README.md
    Start here — what it does and how to install it
  2. 2
    src/__main__.py
    Where the program starts running
  3. 3
    pyproject.toml
    Dependencies and the commands it exposes
  4. 4
    src/__init__.py
    Inside src/ — the main logic begins here
What's in each folder
src/Core code — the actual logic342 files
agentloom-framework-skill/Prompts, skills & agent definitions10 files
docs/Documentation126 files
applications/Folder290 files
test_support/Folder12 files
config/Configuration3 files
patches/Patches2 files
scripts/Helper scripts1 files
READMEHas testsDocumentedCI checksNo licenseUpdated this month
Quick Actions
Details
Creator
linora-u
Language
Python
Category
automation
Published
5/4/2026

Are you the creator of this tool? and earn 85% of every sale.

Show it off in your README:

Featured on WorkflowStacks
[![Featured on WorkflowStacks](https://workflowstacks.com/api/badge/agentloom.svg)](https://workflowstacks.com/skills/agentloom?utm_source=github&utm_medium=badge)
🔥 Hot this week, in your inbox
Skills like this, every Monday.

The five fastest-growing open-source AI skills, ranked by GitHub star growth. One email a week, unsubscribe anytime.

Do this with a working workflow

Pages we wrote and tested that put a skill like this to work — importable n8n templates, outcome guides and Claude Desktop configs.

MCP config

Playwright (browser) MCP config

Drive a real browser for testing and automation from Claude.

Outcome guide

An AI review responder for your store

Paste your reviews into a Google Sheet and the workflow drafts a reply to each in your brand voice. Approve, then post. For high volume with escalation routing, there’s a PRO version.

Free n8n template

Review Reply Drafter

Draft personalized, on-brand replies to every customer review — including careful responses to negative ones. You approve before anything is posted.

Ready-to-run pack · one-time $29

Review Watchdog

Catch every bad review the day it lands — with a reply ready.

Includes the tested workflow and a written setup playbook. Needs: Google Places API key (free tier).

See what you get

Related skills

More automation tools founders pair with this one.

automation★ 208,277
n8n: Scale Your Startup with 207k+ GitHub-Star Automation
n8n lets founders automate workflows across 400+ services, blending visual no‑code design with custom TypeScript code—self‑hosted or cloud.
automation★ 155,897
Langflow: Rapidly Build AI Agents & Workflows
Create and launch AI‑powered agents and multi‑step workflows without deep ML code—ideal for founders scaling generative‑AI products. 155k+ GitHub stars
automation★ 133,722
MoneyPrinterTurbo: AI Video Automation
Generate short videos with one click using AI. For founders needing automation.
automation★ 988
kandev: AI‑Powered Kanban & Dev Environment for Faster Releases
kandev lets founders orchestrate AI agents to manage tasks, review code, and open PRs—all self‑hosted with no telemetry. Ideal for teams seeking automated workflows in Go.
automation★ 51
zoom-ascent: Automate Zoom meetings to boost founder productivity
Zoom-ascent offers ready-to-use HTML scripts that automate attendance, recordings, and transcriptions, letting startup founders run Zoom meetings faster and focus on growth.
automation★ 3
CVE-Intel: Automate CVE Exploit Monitoring
Get automated CVE exploit PoC aggregation for enhanced cybersecurity, ideal for founders seeking proactive threat protection.