cyberneticlibrary
← library

Route agents to local Ollama models

add-ollama-providerskillsetup L2★29,740
nanocoai/nanoclaw ↗
What it does

Route agent group to local Ollama model instance

Best for

When you want to run agent inference locally to cut API costs or experiment with open models.

Inputs
  • · agent group folder
  • · Ollama model name
  • · block-anthropic flag
Outputs
  • · container.json env overrides
  • · shared settings.json with model field
Requires
  • · Ollama daemon on localhost:11434
  • · host Docker
  • · launchctl/systemctl
Preconditions

Ollama installed and running, model pre-pulled, agent group exists

Failure modes
  • · Ollama daemon unreachable
  • · Model not found
  • · Container spawn fails
  • · env var mismatch
Trust signals
  • · 6-step explicit wiring (check source support, configure, set model, build, verify)
  • · Verify script uses curl /api/tags
  • · Covers env, blockedHosts, home perms, launchd reload