cluaizcluaiz/cluaiz-tools/skills/learn-cluaiz/SKILL.md

learn-cluaiz

Skill3 starsChanged 6 days agoApache-2.0View on GitHub

Explains what Cluaiz is and how to use it: installing, running models, the coding agent, tools (skills, plugins, MCP servers), the APIs and CEP, and how to contribute. Use when the user asks about Cluaiz itself, its commands, its API, or how to do something with it.

Reads credentials

In this repository10 other files it gives its agents.
- build-featureSkill
- code-interpreterSkill
- code-reviewerSkill
- deep-researchSkill
- frontend-designSkill
- git-assistantSkill
- githubSkill
- speechSkill
- tool-creatorSkill
- workflow-managementSkill

The file

We split this skill into the rules it gives an agent

1 rule. About 0 are common, 1 is distinctive, and 0 are very specific to this project.

≈0
Common · estimate
Many files like this have a rule of this kind.
≈1
Distinctive · estimate
This author's own choices. Worth a look.
0
Very specific
Tied to their project's names, files or tools. Probably skip.

Common
Distinctive
Very specific

Distinctive:
Do not describe Cluaiz as only for local models: it also serves provider models through the same API.
line 76

≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.

Files

SKILL.md

Learn Cluaiz

Answer plainly and correctly from this skill and its files. If something is not covered, say so instead of guessing, and point to cluaiz --help, cluaiz <command> --help or https://cluaiz.com/docs. Never invent a command, flag, API route or CEP method.

Links such as references/... are files of this skill: read them with read_tool_file (terminal agent: read_item with learn-cluaiz gives the folder, then read_file on folder plus link).

What it is

Cluaiz is an AI engine written in Rust, in pre-release (v0.1.x). One process runs models on the user's own GPU, RAM and CPU and serves them to apps, scripts and agents:

GGUF models through a fork of llama.cpp: chat, vision, audio input, embeddings.
ONNX models through ONNX Runtime: text-to-speech (Kokoro and others), speech-to-text (Whisper), embeddings.
A placement planner that measures free GPU memory and RAM on every load and chooses where the model runs, the context size and the KV cache type. cluaiz show <model> prints the plan.
Models of OpenAI-compatible APIs can be added as providers and used as <provider>/<model>.
APIs: OpenAI-compatible and Ollama-style routes, plus CEP (JSON-RPC 2.0 at POST /v1/cep) for everything Engine-specific.
Tools for models: skills, plugins and MCP servers, with approvals and an audit log.

Quick answers

Answer these directly; read a reference only for more.

Question	Answer
Start the API server	`cluaiz serve`: OpenAI-compatible routes at `http://127.0.0.1:8000/v1`, CEP at `POST /v1/cep`
The token	`Authorization: Bearer <token>`, the token in `~/.cluaiz/session.token` (when the owner turned authentication on), or an API key from `tokens.create`
List installed tools (CEP)	`tools.list`; install with `tools.install`, remove with `tools.remove`, call a function with `tools.call`
All CEP methods	`rpc.discover`
Chat with a model	`cluaiz run <model>`; the coding agent: `cluaiz agent`
Where a model would run	`cluaiz show <model>`

Which file to read

The user asks about	Read
Installing, first model, chat, serving	references/getting-started.md
`cluaiz agent`, `.agents/`, rules, workflows, memory	references/cli-and-agent.md
How the Engine works inside, the repository folders	references/engine-architecture.md
The HTTP API, CEP methods, tokens	references/cep-api-reference.md
Reporting bugs, contributing, building from source	references/contributing-guide.md
Something none of these cover	references/dynamic-retrieval.md

Tools

Type	What it is	Needs on the machine
Skill	Instructions the model reads when a task fits (`SKILL.md`); some also offer functions run by a script	Nothing, or Python for skills with functions
Plugin	A program the Engine runs, usually one WebAssembly file in a sandbox	Nothing, for WebAssembly
MCP server	An outside program using the Model Context Protocol	Usually Node or `uv`
Extension	A mini-application in the App	In development

The model first sees only short entries for the installed tools. When a tool fits, it loads that tool's instructions and functions, then makes ordinary tool calls. The Engine runs each call, asks for approval when needed, and returns the result.

Kind	In the catalog
Plugins	`math`, `text`, `time`, `rank`, `web-search`
Skills	`code-interpreter`, `code-reviewer`, `git-assistant`, `github`, `build-feature`, `frontend-design`, `deep-research`, `speech`, `tool-creator`, `workflow-management`, `learn-cluaiz`
MCP servers	`browser`, `documents`, `filesystem`, `git`, `fetch`, `context7`, `everything`

cluaiz skill install code-reviewer
cluaiz plugin install web-search
cluaiz mcp install filesystem
cluaiz plugin config web-search TAVILY_API_KEY
cluaiz skill list

How to answer

Give the exact command or request, then one sentence on what it does.
Say "pre-release" when the user may rely on something staying the same: settings, APIs and file layout can still change.
Do not describe Cluaiz as only for local models: it also serves provider models through the same API.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

It worked
Partly
It didn't

What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

Check your setup, free