Our take
Activeregular commits and releases- Activity
- 62
- Community
- 65
- Issues
- 90
- Pull requests
- 52
+6 stars in the last 30 days
- Closes issues faster than 82% of the projects we track
Graphon is an actively maintained, still-evolving execution engine with a broad contributor base and frequent releases, making it promising for teams that want to build agent workflows in Python and are comfortable evaluating an evolving project.
Good fit if
- You want graph-based execution and explicit workflow state for agentic AI workflows.
- Your stack can use Python 3.12 or 3.13, and you want built-in nodes plus integration protocols to extend.
- You want an Apache-2.0 licensed project with recent commits and releases and contributions from multiple people.
Look elsewhere if
- You need a turnkey agent application rather than an engine to build on.
- Your runtime cannot use Python 3.12 or 3.13.
- You require a long-established, settled workflow engine rather than one the project describes as still evolving.
All health signals
| Last commit | 2026-11-02 (6 days ago) |
|---|---|
| Commits, last 90 days | 37 |
| Releases, last 12 months | 12 (latest v0.8.0, 2026-09-09) |
| Contributors | 19 (top contributor: 34% of commits) |
| Issues closed, last 90 days | 28 (typically closed in 0 days) |
| Pull requests merged, last 90 days | 37 (typically merged in 1 day) |
| Project age | 8 months |
Checked on 2026-11-09 with the GitHub API.
Project history
Daily GitHub snapshots since Oct 7, 2026. History covers stars and forks; health signals are not tracked historically.
Stars
90Forks
47Change since Nov 16, 2026: +1 stars·+1 forks
Overview
Graphon is a Python graph execution engine for agentic AI workflows. It lets developers define workflows as graphs, validate and build them, then run them through a queue-based, event-driven engine with shared workflow state.
The repository also includes built-in workflow nodes and integration protocols for HTTP, files, tools, and human input. Its DSL import path includes Slim-backed LLM nodes. Graphon is aimed at developers building or extending agent workflows, rather than end users looking for a ready-made assistant.
Key Features
- Queue-based engine orchestration with event-driven execution.
- Graph parsing, validation, and fluent graph construction.
- Frame-owned runtime state and variables alongside shared workflow state.
- Built-in nodes for common workflow patterns.
- DSL import support, including Slim-backed LLM nodes.
- Integration protocols for HTTP, files, tools, and human input.
- Extensible engine layers and external command channels.
- An end-to-end Slim LLM example with DSL import and direct Python graph construction.
Use Cases
- An AI application developer can compose agent steps into a graph when execution order and workflow state need to be explicit.
- A platform team can add custom engine layers or external command channels to adapt workflow execution to a host application.
- A team prototyping LLM workflows can use the Slim example to compare DSL import with direct Python graph construction.
- Developers integrating agents with outside systems can use the provided HTTP, file, tool, and human-input protocols as workflow integration points.
What you need
Detected in the repository
- Python >=3.12,<3.14 (from pyproject.toml)
- A test suite and automated checks on GitHub Actions
License in plain words
Apache-2.0permissive
- Commercial use: yes
- Modify and redistribute: yes
- You must keep: the license, the NOTICE file and a note of your changes
- Share your changes: no
- Includes an explicit patent grant from the contributors:
A summary, not legal advice: the LICENSE file is what applies.
Getting Started
Follow the development setup, then run the Slim LLM example. The example covers both DSL import and direct Python graph construction. See the README for project documentation links.
Alternatives
- mcp-agent: mcp-agent centers on MCP server integrations and composable multi-agent workflows, while Graphon provides a general Python graph execution engine.
- AgentRun: AgentRun is a self-hosted platform for managing agent revisions and durable runs, while Graphon focuses on executing graph-defined workflows.
| Project | Language | License | Stars | Status |
|---|---|---|---|---|
| Graphon | Python | Apache-2.0 | 86 | Active |
| mcp-agent | Python | Apache-2.0 | 8.6k | Quiet |
| AgentRun | TypeScript | Apache-2.0 | 423 | New |
Considerations
Graphon is still evolving, so teams should validate its graph execution and integration surfaces against their own workflows before depending on them. It requires Python 3.12 or later in the 3.12 series, or Python 3.13, which may constrain existing environments. The project reports active development, with regular commits and releases, issue closures, and pull request merges; contributors are not concentrated around one person. The repository lists no Docker, Compose, or Helm requirement. It is licensed under Apache-2.0.
Found this useful?
Share it with someone who would like Graphon.