Manage Ollama task queue and code generation on a remote worker machine via SSH.
Provides queue health checks and task execution with exclusivity locking.
ollama status [clean] [--force] [--kill-ollama]Check queue status and Ollama server health.
clean: remove stale locks (prompts if Ollama is busy)clean --force: remove stale locks without promptingclean --kill-ollama: stop and restart Ollama server to cancel active generationollama run Execute a task on the worker machine. Available tasks:
| Task | Usage | Description |
|---|---|---|
| --- | --- | --- |
ping | ollama run ping | Check runner is alive |
codegen | ollama run codegen | Generate code via Ollama |
write | ollama run write | Generate and write code to file |
test | ollama run test [suite] | Run project tests |
exec | ollama run exec | Run arbitrary shell command |
list-projects | ollama run list-projects | List available projects |
| Variable | Default | Description |
|---|---|---|
| --- | --- | --- |
OLLAMA_WORKER_HOST | worker-mac | SSH host alias |
OLLAMA_RUNNER_PATH | ~/worker/runner | Runner scripts path on worker |
WORKER_ROOT | $HOME/worker | Worker root directory |
PROJECTS_DIR | $WORKER_ROOT/projects | Projects directory |
DEFAULT_PROJECT | _(none)_ | Default project for context |
OLLAMA_MODEL | qwen2.5-coder:32b | Ollama model |
OLLAMA_URL | http://localhost:11434/api/generate | Ollama API endpoint |
skill.py OpenClaw skill entry point
runner/
queue_status.sh Queue and server health checker
run_task.sh Task executor with Ollama integration
README.md Setup and usage guide
SKILL.md This file
共 1 个版本