enable-ollama

name	enable-ollama
description	Verifies and sets up local Ollama — checks install, starts daemon, pulls model, verifies CORS and OpenAI-compatible endpoint. Client-agnostic; use before any skill that needs a working Ollama (e.g. enable-ollama-in-cursor). Use when the user wants Ollama running locally, needs to troubleshoot Ollama, or before configuring any client to use Ollama.

Enable Ollama

Ensures Ollama is installed, running, has the requested model pulled, CORS is configured, and the OpenAI-compatible /v1 endpoint responds. This skill is client-agnostic — it knows nothing about Cursor, VS Code, or any other consumer.

Defaults

Item	Default	Notes
OLLAMA_CURSOR_MODEL	gemma4:26b	Override with any ollama list name.
OLLAMA_HOST	http://127.0.0.1:11434	Override if Ollama binds elsewhere.
OLLAMA_ORIGINS	*	Required for browser/ngrok clients.

Agent workflow (run in order)

1. Quick health check (recommended first)

Runs all checks in sequence:

bash ~/.agents/skills/enable-ollama/scripts/health-check.sh

If this passes, Ollama is ready. Skip to step 5 for endpoint URL.

2. Individual checks (if health check fails)

Each script exits 0 on success, non-zero on failure:

# Check ollama CLI installed
bash ~/.agents/skills/enable-ollama/scripts/check-install.sh

# Check daemon responding
bash ~/.agents/skills/enable-ollama/scripts/check-daemon.sh

# Pull/verify model
bash ~/.agents/skills/enable-ollama/scripts/pull-model.sh

# Check CORS configuration
bash ~/.agents/skills/enable-ollama/scripts/check-cors.sh

3. Fix CORS (if check-cors.sh warns)

export OLLAMA_ORIGINS='*'
# Then restart Ollama from this shell, or on macOS:
launchctl setenv OLLAMA_ORIGINS "*"

4. Re-run health check after fixes

bash ~/.agents/skills/enable-ollama/scripts/health-check.sh

5. Result: endpoint URL

After health check passes, the Ollama OpenAI-compatible endpoint is:

${OLLAMA_HOST:-http://127.0.0.1:11434}/v1

Use this URL in any OpenAI-compatible client (Cursor, Continue, etc.).

Script reference

Script	Purpose	Destructive?
check-install.sh	Verify ollama CLI exists	No
check-daemon.sh	Verify :11434 responding, auto-start macOS	Yes (start)
pull-model.sh	Pull model if missing	Yes (pull)
check-cors.sh	Check OLLAMA_ORIGINS=*	No
health-check.sh	End-to-end check (calls all above)	Yes (daemon+pull)

All scripts are idempotent — safe to run repeatedly.