Ten minutes to your first publishable draft.
GoTo Agents is a local Python + FastAPI app. You need Python 3.10+ and an API key for at least one model provider (Anthropic, OpenAI, or Gemini). One key is enough for everything on this page.
Install and start
git clone https://github.com/staceyfarias/goto-agents.git
cd goto-agents
pip install -r requirements.txt
python start.py
Open http://localhost:8000/app in your browser — the app binds
to localhost only. Add your provider key under
Settings → LLM Configs. A
GETTING_STARTED.md with this same walkthrough is seeded into
your instance directory.
Import the demo: Pulse Studio
Menu → Import Workspace → choose
examples/pulse-studio.zip from the install directory. You get a
three-agent content studio: Zita researches,
Quill writes, Portia edits. Open a chat
with Zita and run the loop on a topic you actually care about:
Scan current trends in <your topic> and suggest 3 directions for a LinkedIn post.Deep dive on direction 2 and build a content pack.- Switch to Quill:
Draft a LinkedIn post from Zita's content pack. Goals: practitioners, conversational, ~300 words. - Switch to Portia:
Review the draft against my goals and the research.
Loop Quill ↔ Portia until Portia says ready — the output is a
real post. The workspace's documents/demo-script.md is the
guided tour, including the show-off moves: inter-agent consultation,
@-mentioning the style guide, memory that survives sessions, and the
rendered-prompt view where you can read the exact system prompt any agent
gets.
Where to next
That was the tour. When you're ready to build a workspace for your work — by hand, or with a Bootstrap Agent that designs it with you — head to Getting Started. For the mental model behind what you just saw, read the core concepts.