Auto-GPT is the project that put the autonomous-agent idea on the map in 2023. It runs an LLM in a continuous loop, lets it pick tools, and chases an arbitrary high-level goal. The codebase is Python, the configuration is split between code and JSON, and the agent runs in a single-process command-line setup.
Spawn explorers, fact-checkers, and specialists from a coordinator with one tool call. No fork required.
runtime.max_turns is a hard ceiling, set once in the workspace. The agent stops cleanly instead of spending indefinitely.
No env-var scavenger hunts. Tools, brains, prompts, triggers, and budgets are visible in a single config.
Open Studio and tell the builder assistant what you want. It builds the agent for you, connects your tools, and you refine it live on the canvas. It's vibe coding for agents, nothing to install.
Open StudioEngineering notes from the Digitorn team. No marketing, no launch announcements, no "10 prompts that will change your life". Just the things we write that we'd want to read.
