Complex workflows. Obvious code.
Sley turns tangled workflow logic into code your team can scan and change with confidence. A workflow rarely starts tangled. It gets there one condition, callback, counter, and fallback at a time, until the work still makes sense but its shape no longer does.
Sley puts that shape back in front of you. Write ordinary functions, connect the paths they may take, and run the graph. Branching, fan-out, joins, retries, and nested workflows stay visible, the final state comes back directly, and your application remains ordinary Python or TypeScript.
- A node is one ordinary synchronous or asynchronous function.
- A link makes one allowed next step visible.
emit()chooses one or more paths.- A Flow waits for those paths and
run()returns their final shared state.
Use Sley when branching and synchronization have made a workflow's shape hard
to see. Keep ordinary calls, conditions, loops, asyncio.gather, or
Promise.all while they still explain the workflow clearly.
pip install sleynpm install @jigging/sleyPython requires version 3.13 or newer. Package and runtime details live in the Python and TypeScript references.
The Quickstart takes one file from installation to visible output. The Learn path then evolves that same release workflow from a linear graph through routing, branch data, fan-out, joins, nested boundaries, failure evidence, and advanced graph design.
Sley evolved through PocketFlow and Caskada. The lineage records what changed and which tradeoffs remain intentional.
Sley is licensed under the Mozilla Public License 2.0.