Skip to content

usage.md

Latest commit

 

History

History
126 lines (91 loc) · 3.42 KB

usage.md

File metadata and controls

126 lines (91 loc) · 3.42 KB
Raw

Usage

Most people use the CLI. Some teams prefer the web UI. The API is for wiring it into your own tooling. All three call the same generator.

CLI — the short version

agent-generator "<your sentence>" -f <framework> [-o <path>]

That's the only shape you need to remember.

Recipes

Generate a CrewAI research team

agent-generator "Research team with a researcher and a writer" \
  -f crewai -o research_team/

You get a full CrewAI project — crew.py, agents.yaml, tasks.yaml, tool stubs, tests, pyproject.toml, README.

Export a WatsonX Orchestrate agent

agent-generator "Customer support assistant that triages tickets" \
  -f watsonx_orchestrate -o support.yaml

orchestrate agents import -f support.yaml

Build a LangGraph data pipeline on OpenAI

agent-generator "Pipeline that extracts, transforms, loads data" \
  -f langgraph -p openai --model gpt-4o -o pipeline.py

Wrap any Python output as an MCP server

agent-generator "Research assistant" -f crewai --mcp -o assistant.py
python assistant.py          # FastAPI on :8080, POST /invoke

Estimate cost before generating

agent-generator "Multi-step content workflow" -f crewai_flow --show-cost

Try without credentials

agent-generator "Anything" -f react --dry-run

All the CLI flags

Flag Default What it does
-f, --framework (required) crewai, crewai_flow, langgraph, react, watsonx_orchestrate
-p, --provider watsonx watsonx or openai
--model provider default Override the model.
--mcp / --no-mcp off Add a FastAPI /invoke wrapper to Python output.
-o, --output stdout Write to a file or directory.
--dry-run off Skip the LLM call (fake spec, real renderer).
--show-cost off Print token + cost estimate before generating.
--temperature 0.7 Sampling temperature for the planning call.
--max-tokens 4096 Cap on the planning response.

Web UI

uvicorn agent_generator.wsgi:app --host 0.0.0.0 --port 8000 --reload

Open http://localhost:8000 and you get:

  1. A text area to describe the agent team.
  2. A framework picker (or Auto).
  3. An artifact-mode picker (code, YAML, or both).
  4. A tool catalogue (six prebuilt templates).
  5. A file tree + syntax-highlighted preview after generation.
  6. A Download ZIP button.

The full platform (make start) adds auth, projects, runs, marketplace, and audit on top of the same screens.

REST API

Path Method What it does
/api/plan POST Run the planner; get a ProjectSpec JSON.
/api/build POST Render files from a ProjectSpec.
/api/generate POST Plan + build in one call.
/download/{id} GET Download a generated bundle as ZIP.
/health GET Liveness probe. 
curl -X POST http://localhost:8000/api/generate \
  -H 'content-type: application/json' \
  -d '{"prompt": "Research team", "framework": "crewai", "provider": "watsonx"}'

Docker

docker build -t agent-generator .
docker run -e WATSONX_API_KEY=... -e WATSONX_PROJECT_ID=... \
           -p 8000:8000 agent-generator

CLI inside the container:

docker run --rm -e WATSONX_API_KEY=... \
  agent-generator agent-generator "Say hi" -f react --dry-run

Next: Pick a framework · Architecture