AutoKitteh

Normal
GitHub Go Apache-2.0

Description

AutoKitteh is a durable workflow automation platform for AI agents and integrations, with code-first workflows, long-running state, and deep observability.

Key Features

  • Durable execution engine that survives crashes, restarts, and multi-day schedules
  • Code-first workflows written in Python with versioning and type checking
  • Native integration adapters for GitHub, Slack, Google Workspace, HTTP, and databases
  • Webhook and polling triggers with idempotent retry and replay support
  • Built-in scheduler for cron-style and event-driven automations
  • {"First-class observability":"timeline, logs, and replayable step history"}

Use Cases

💡 Orchestrating AI agent runs that span multiple API calls over hours or days
💡 Building production incident-response automations with retry and human approval
💡 Replacing fragile cron scripts with durable, observable workflows
💡 Driving PR review pipelines that connect GitHub, Slack, and AI code reviewers
💡 Synchronizing customer data between SaaS tools without a heavyweight iPaaS

Tags

workflow durable-execution automation agent-platform

Categories

🔄 Workflow 🤖 Agent Framework

Quick Start

pip install autokitteh
# workflow.py
from autokitteh import trigger, workflow
@workflow
async def on_pr_opened(event):
    await event.slack.post("#code-review", f"PR opened: {event.pr.url}")
auto = autokitteh.workflow("workflow.py")
auto.deploy()
Visit GitHub Visit Website View Docs

Related Projects