porto is two long-running services plus a local Postgres. The pins below are what CI runs. Confirm on your machine:
A green check means your toolchain matches CI.

Optional, for CLI seats

Only needed if you will use those providers: Details: CLI harnesses.

Optional, for local models

  • Ollama if you want the Ollama provider
  • Any OpenAI-compatible server (vLLM, LM Studio, llama.cpp, a Tailscale box) if you want Local / network
Details: Local models.