Agent skill · software engineering · mvanhorn
pp-ollama-cloud
Routes every prompt to the right hosted Ollama model. Wraps chat, embeddings, catalog, and the OpenAI-compatible... Trigger phrases: `pick an ollama model`, `advise on a model`, `which ollama cloud model should I use`, `route this prompt`, `use ollama cloud`, `run ollama cloud`.
Why this skill is useful
Adds executable commands for intelligent model routing and comparison that extend the AI's capabilities in using the Ollama Cloud API.
What it needs
Requires ollama-cloud-pp-cli installed locally. About 6k tokens when loaded. Last updated 2026-08-06. 1,890 stars on the source repository.
What this skill does
Ollama Cloud — Printing Press CLI Prerequisites: Install the CLI This skill drives the ollama-cloud-pp-cli binary. You must verify the CLI is installed before invoking any command from this skill. If it is missing, install it first: 1. Install via the Printing Press installer. It defaults binaries to $HOME/.local/bin on macOS/Linux and %LOCALAPPDATA%\Programs\PrintingPress\bin on Windows: 2. Verify: ollama-cloud-pp-cli --version 3. Ensure the reported install directory is on $PATH for the agent/runtime that will invoke this skill. If the npx install fails (no Node, offline, etc.), fall back to a direct Go install (requires Go 1.26.6 or newer): If --version reports "command not found" after install, the runtime cannot see the binary directory on $PATH. Do not proceed with skill commands until verification succeeds. When to Use This CLI Pick this CLI when you need agent-driven access to Ollama's hosted catalog and want routing intelligence baked in. Especially useful when the alternative is hardcoding a model name and hoping it stays right. Pairs naturally with multi-backend setups where ollama-cloud is one of several providers. Unique Capabilities These capabilities aren't available in any other tool for this API. Routing intelligence advise — Picks the right Ollama Cloud model for a prompt by combining live catalog, heuristic prompt-feature extraction, curated cost/latency metadata, and an optional cheap meta-LLM tiebreak. When an agent needs to pick a hosted Ollama model and the default routing is wrong, reach for advise instead of hardcoding the model name. compare — Runs the same prompt against N hosted models in parallel and emits side-by-side response, tokens, and latency. Use when calibrating advisor recommendations or picking between two close models. advise — With --explain, advise emits the full scoring trace: feature extraction, per-model scores, filter passes, tiebreak rationale. …
How to use it
Reference it in AdaL, Claude Code, Cursor or any coding agent — nothing to install:
@skills mvanhorn/ollama-cloud