Define Operations, compose Pipelines, let AI Agents run the work end-to-end. Code quality automation is built in as a plugin.
✓ lint-check · 2.1s
✓ type-check · 4.8s
▸ ai-review · running
Capabilities
Build, orchestrate, and execute complex AI automation workflows with a small set of composable core primitives.
Define typed inputs for your pipelines — folders, code files, GitHub projects, or custom types via plugins.
Define atomic coding tasks with configurable executor backends — AI Agents (Claude, Codex, Gemini) or scripts.
Chain Operations into multi-step DAG workflows with typed nodes, conditional branching, and parallel execution.
Pluggable AI Agent capabilities. Skills decouple 'what the agent knows' from 'what the operation does' — both independently reusable.
Define when and how pipelines are triggered automatically. Connect events (code changes, schedules) to pipeline execution.
Track background pipeline execution with real-time progress, execution traces, structured output, and timing metadata.
How it works
Break down complex tasks into atomic operations, configure AI Agents or scripts as executor backends with typed inputs and outputs.
Connect Operations into DAG workflows with conditional branching, parallel execution, and typed nodes and edges.
Inject reusable Skills into Agents, define trigger conditions via Rules — code changes, schedules, or manual triggers.
Start a Job to execute your Pipeline in the background. Track progress in real-time, view traces and structured output.
Open Source
Ordine is fully open source. Inspect every line, contribute improvements, and build on a foundation you can trust.
Free to use, modify, and distribute for any purpose.
Built in the open with contributions from developers worldwide.
Extend with plugins — custom Operations, Skills, and domain logic.
Run entirely on your own infrastructure with full data ownership.
FAQ
Ordine is an AI Agent first work orchestration framework. It provides universal workflow orchestration — define Operations, compose Pipelines, let AI Agents run the work end-to-end. Code quality automation is built in as a plugin.
Ordine focuses on higher-level work orchestration. LangChain helps chain LLM calls, while Ordine manages entire workflow lifecycles — including a typed DAG execution engine, plugin system, event-driven rule triggers, and background Job tracking.
Yes. Ordine is fully self-hostable with PostgreSQL as the database. Docker and docker-compose files are provided for easy deployment.
Ordine Operations support any AI Agent as executor — Claude, Codex, Gemini, or your own custom Agent. Script execution mode is also supported.
Ordine is currently in Preview stage — APIs and features may change. The core engine is stable, with new features and improvements released regularly. Feedback and contributions are welcome.