OSS Tanbou
View on GitHubProject website

combine role-based Crews with event-driven Flows to build multi-agent automation

OSS scale score 321.7OSS health 93Update pace 107 commits/30d
About these scores

OSS scale score is an unbounded metric that log-compresses and weights Stars, Watchers, Forks, and Contributors. Discovery score is the current OSS scale score minus the score at discovery. Update pace is commits in the last 30 days, growth momentum is the OSS scale score difference within the recent observation window, and OSS health is a 0–100 rating based on available recency, Community Health, and release data.

Stars
59,327
Primary language
Python
License
MIT
Repository last updated
Oct 3, 2026
On this page

Overview

CrewAI is a Python framework for production-oriented multi-agent workflows. Crews coordinate role-based agents and tasks, while Flows provide event-driven execution, state, and branching. The two can be combined so autonomous collaboration and explicit workflow control live in the same application.

Features and best fit

Based on official documentation; not hands-on tested · Content checked:

Key features

Coordinate role-based agents and tasks with Crews

Agents can be configured with roles, goals, tools, and LLMs, then assigned tasks under sequential or hierarchical processes. The JSON-first scaffold separates agent definitions from crew-level workflow configuration.

Sources: [1]

Control state and branching with event-driven Flows

Flows add state management, event-driven execution, and conditional branching so Crews or individual LLM calls can be composed with normal Python application logic.

Sources: [1]

Best fit

Fits systems that need both agent autonomy and explicit workflow control

CrewAI is useful when research, analysis, or generation is delegated across agents but important sequencing, state transitions, and branches still need deterministic application-level control.

Sources: [1]

Before adoption

Design model credentials, tool permissions, cost controls, and output validation separately

Runs require credentials for the selected model provider, and tool integrations may require additional credentials. Production deployments should define permission boundaries, retries, human review, cost limits, and output validation around agent execution.

Sources: [1]

CrewAI 1.15.23 requires Python 3.10 through 3.13

The 1.15.23 crewai package metadata requires Python >=3.10,<3.14 and pins the matching core and CLI components to 1.15.23.

Sources: [2][3]

Official sources

  1. [1]CrewAI 1.15.23 README(2026-10-03)
  2. [2]CrewAI 1.15.23 package metadata(2026-10-03)
  3. [3]CrewAI 1.15.23 release(2026-10-03)
  4. [4]CrewAI MIT license(2026-10-03)
Supplemental curator note

CrewAI combines autonomous role-based collaboration with explicit stateful workflow control. Production use still requires careful handling of model credentials, costs, tool permissions, output validation, and failure behavior.

Try it in 3 steps

  1. 1

    Install the CrewAI 1.15.23 CLI

    Pin the CLI on a Python >=3.10,<3.14 environment.

    uv tool install "crewai==1.15.23"
  2. 2

    Scaffold a JSON-first crew project

    Create a project with agents/*.jsonc and crew.jsonc so roles, tasks, and the process can be configured explicitly.

    crewai create crew quickstart
  3. 3

    Configure provider credentials and run the crew

    Set the selected model provider API key in .env first. Review external model costs and the tools/actions granted to agents before execution.

    cd quickstart && crewai install && crewai run
Check the official README

Growth

Growth trends · Last 30 days

59,327 Stars

Trend data is still being collected.

Development activity

Last 90 days · weekly

Commits (last 30 days)
107
Open PRs
323

Development activity is still being collected.

Built with

Categories and tags

Categories

Tags

GitHub data

GitHub dataView detailed GitHub data
View on GitHub

GitHub Topics

  • agents
  • ai
  • ai-agents
  • llms
  • aiagentframework
Stars
59,327
Forks
8,633
Watchers
398
Open issues
211
Contributors
336
Owner type
Organization
Primary language
Python
License
MIT
Repository last updated
Oct 3, 2026
Write a related article

Share a guide or use case for this OSS in Markdown. Articles are published after administrator approval.

Report incorrect information

Tell us if any listing information is incorrect or outdated.

After reading this page, do you know what to do next?