Back to Home
· product · apflow
apflow
Distributed task orchestration built on apcore
apflow 0.22.1 is a Python task-orchestration engine that combines structure trees for organization with dependency DAGs for execution order. It supports durable execution, scheduling, SQLite or PostgreSQL storage, and distributed workers. Its registered operations can be exposed through apcore adapters, while apflow retains its own orchestration model, compatibility, and release lifecycle.
· 01
Features
01Dual Orchestration Model: Structure trees organize work while dependency DAGs control execution order
02Task Reuse Modes: Create, link, copy, archive, and mixed workflow construction
03Durable Execution: Checkpoint and resume, retry with backoff, and circuit-breaker support
04Cost Governance: Token budgets, policies, model downgrade chains, and usage reporting
05Scheduling: Poll-based schedules, webhook triggers, and run history
06Distributed Workers: Leader election, task leasing, and PostgreSQL coordination
07Storage: SQLite for local operation and PostgreSQL for distributed deployments
08Protocol Surfaces: REST, MCP, A2A, and CLI entry points over registered apcore modules
· 02
Get Started
Python · Implementation
Python 3.11+ implementation of the apflow orchestration engine.
Install
$ pip install apflowQuick Start
python.pyfrom apflow.app import create_app
app = create_app()
# The registry contains apflow operations that can be
# called directly or projected through an apcore adapter.
modules = list(app.registry.list())· related
