Autonomous AI Agency
A self-hosted multi-agent platform that audits a website's technology stack, then provisions a fleet of specialist AI agents to plan, execute and verify work against that infrastructure, with automatic model-provider failover.
Autonomous AI Agency starts by scanning a given website URL against more than 1,000 recognized technology stacks to produce a technical audit, then provisions a fleet of industry-specific specialist agents that plan, execute and verify further work under a human approval gate, running continuously on the operator's own hardware with no data leaving their network.
A lead agent orchestrates the specialist fleet with automatic provider failover. The project reports that a large share of its own merged pull requests were opened by its own agent fleet, and publishes its self-audit output as evidence of what the system produces.
A local run needs no cloud account, database server or Docker; it stores data in SQLite and launches the dashboard and its 24-hour autonomy loops from a single process, started with one script. It is written in Python, requires Python 3.11 and Node 20, and is MIT licensed.