ov-images/skills/openclaw-ollama/SKILL.md
Headless OpenClaw gateway with local Ollama LLM inference. GPU-accelerated, no desktop. Use when working with the headless openclaw+ollama deployment MUST be invoked before building, deploying, configuring, or troubleshooting the openclaw-ollama image.
npx skillsauth add overthinkos/overthink-plugins openclaw-ollamaInstall this skill globally with one command. Works with Claude Code, Cursor, and Windsurf.
3 of 9 scanners reported clean
Some scanners were skipped, did not run, or reported a non-clean status. Review each row below.
Headless OpenClaw gateway with local LLM inference via Ollama — no desktop, no browser.
| Property | Value | |----------|-------| | Base | nvidia | | Layers | agent-forwarding, openclaw, ollama | | Platforms | linux/amd64 | | Ports | 18789, 11434 | | Registry | ghcr.io/overthinkos |
fedora → nvidia (CUDA base)pixi → python → supervisord (transitive)nodejs (transitive)openclaw — gateway on :18789ollama — LLM server on :11434| Port | Service | Protocol | |------|---------|----------| | 18789 | OpenClaw gateway + Control UI | HTTP | | 11434 | Ollama API | HTTP |
ov image build openclaw-ollama
ov config openclaw-ollama
ov start openclaw-ollama
# Pull a model and configure OpenClaw to use it
ov shell openclaw-ollama -c "ollama pull llama3"
/ov-layers:openclaw — AI gateway service/ov-layers:ollama — LLM inference server/ov-images:openclaw — gateway only (no LLM, no GPU)/ov-images:ollama — LLM only (no gateway)/ov-images:openclaw-ollama-sway-browser — adds desktop + ChromeAfter ov start:
ov status openclaw-ollama — container runningov service status openclaw-ollama — all services RUNNINGcurl -s http://localhost:18789 — OpenClaw gateway respondscurl -s http://localhost:11434/api/tags — Ollama API respondsOpenClaw gateway (18789) uses port relay (socat) — the gateway binds to loopback, socat forwards from the container interface. This avoids the allowedOrigins requirement for the Control UI.
MUST be invoked when the task involves the openclaw-ollama image, running OpenClaw with local models, or headless openclaw+ollama deployment. Invoke this skill BEFORE reading source code or launching Explore agents.
/ov:image — image family umbrella (image: entries in overthink.yml, build/validate/inspect/list)/ov:build — build.yml vocabulary (distros, builders, init-systems)tools
Use when authoring or modifying a charly PLUGIN — a candy with a `plugin:` block that contributes Providers (verbs/kinds/deploy-targets/steps/builders/commands), its own CUE schema, builtin (compiled-in) or external (out-of-tree git repo). Covers the unified Provider model, the per-plugin CUE-schema contract (single source → Go params for dev + schema-over-Describe RPC for runtime), the SDK, and the loader.
tools
The CUE data-validation / configuration CLI (cue), pinned to v0.16.1. Use when working with the cue candy, installing the cue binary into a box or onto a target:local dev host, or running the offline schema-vendoring pipeline that feeds charly's egress validation.
tools
CUE EGRESS validation — validating (and, where it adds value, generating) the config files charly WRITES to a system BEFORE the bytes hit disk. MUST be invoked before working on charly/egress.go, the vendored schemas under candy/plugin-egress/egress-schemas/vendor/, the ValidateEgress / registerVendoredEgressKind path, the offline `task cue:vendor` pipeline, or adding an egress schema for any written artifact (cloud-init, k8s manifests, traefik routes, runtime config, install ledger, systemd/quadlet units, ssh_config, libvirt XML).
tools
Kubernetes cluster-probe declarative check verb — the `kube:` check verb (nodes, pods, ingress, storage class, addon health, apply/delete, and arbitrary resource GETs) served out-of-process by the candy/plugin-kube plugin (vendored client-go; no external kubectl required).