chat-format — Claude Code skill: what it does, how to install
Formats chat messages into local model templates like llama3 or chatml for Ollama and llama.cpp, and routes queries across intents using a compact WASM HNSW index.
npx skills add ruvnet/ruflo --skill chat-format -a claude-codeWhat chat-format does
The skill formats chat message arrays containing roles and contents into prompt strings for local runtimes like llama.cpp and Ollama. Supported template options include llama3, mistral, chatml, phi, gemma, or an automatic match by model id. It does not send queries to models or format strings for Anthropic, OpenAI, or Gemini, because hosted APIs require structured objects rather than raw prompt templates.
A secondary capability is an in-memory WASM pattern router built on HNSW. The ruvllm-wasm 2.0.1 build holds around 11 patterns at most, which makes it suitable for routing between named intents rather than document search. Router instances exist solely within the running MCP server process and reset when that process restarts.
Running this tool requires @ruvector/ruvllm-wasm installed via npm i -g @ruvector/ruvllm-wasm. If the module is missing, formatting calls fail immediately.
What it helps with
- format chat arrays into raw prompt strings for llama3 or mistral
- prepare dialog history for local execution in Ollama or llama.cpp
- route user queries between up to 11 intent patterns via WASM HNSW
- verify ruvllm WASM module availability using ruvllm_status
Example prompts for the agent
- Format this conversation history using the llama3 template for local inference.
- Convert these message objects into a raw chatml prompt string.
- Create an HNSW router with 384 dimensions and route the query embedding.
How to install chat-format
1Skills CLI (Vercel)
npx skills add ruvnet/ruflo --skill chat-format -a claude-code2Manually: put the files in place
~/.claude/skills/chat-format/SKILL.mdCopy the whole skill folder; for one project use .claude/skills/chat-format/
1Skills CLI (Vercel)
npx skills add ruvnet/ruflo --skill chat-format -a codex2Manually: put the files in place
.agents/skills/chat-format/SKILL.mdFor all projects use ~/.agents/skills/
1Skills CLI (Vercel)
npx skills add ruvnet/ruflo --skill chat-format -a cursor2Manually: put the files in place
.cursor/skills/chat-format/SKILL.mdCursor also reads .agents/skills/ and .claude/skills/
1GitHub CLI
gh skill install ruvnet/ruflo chat-format2Manually: put the files in place
.github/skills/chat-format/SKILL.md1Skills CLI (Vercel)
npx skills add ruvnet/ruflo --skill chat-format -a gemini-cli2Platform CLI
gemini skills install https://github.com/ruvnet/ruflo --path plugins/ruflo-ruvllm/skills/chat-formatManually: put the files in place
.devin/skills/chat-format/SKILL.mdThe legacy .windsurf/skills/ path still works
Skills CLI (Vercel)
npx skills add ruvnet/ruflo --skill chat-format -a opencodeClaude.ai
Zip the folder and upload it in Settings → Capabilities → Skills
From the same set104
The whole set →Similar
Questions about chat-format
How to install chat-format in Claude Code?
Run npx skills add ruvnet/ruflo --skill chat-format -a claude-code.
What does chat-format do?
It converts chat messages into prompt strings for local models like chatml or llama3 and routes intents through an in-memory WASM HNSW index.
What dependencies are required for chat-format?
You need @ruvector/ruvllm-wasm installed via npm i -g @ruvector/ruvllm-wasm. Without this package, MCP tool calls will fail.
Source text
SKILL.md · MIT
---
name: chat-format
description: Format chat messages in a local model's template (llama3, mistral, chatml, phi, gemma, or a model id) and route queries over a small WASM HNSW pattern index (about 11 patterns). Use when preparing prompts for local runtimes.
argument-hint: "<prompt> [--provider anthropic|openai|gemini|ollama|cohere]"
allowed-tools: mcp__plugin_ruflo-core_ruflo__ruvllm_chat_format mcp__plugin_ruflo-core_ruflo__ruvllm_hnsw_create mcp__plugin_ruflo-core_ruflo__ruvllm_hnsw_add mcp__plugin_ruflo-core_ruflo__ruvllm_hnsw_route mcp__plugin_ruflo-core_ruflo__ruvllm_status Bash
---
# Chat Format
Format prompts for multi-provider LLM inference with context retrieval.
## When to use
When you need a prompt string in a local model's chat template (for llama.cpp, Ollama or another local runtime), or a small embedding-based router over a handful of patterns. It does not call a model and it does not format for hosted APIs: Anthropic, OpenAI and Gemini take structured messages, not a template string.
## Steps
1. **Format chat** — call `mcp__plugin_ruflo-core_ruflo__ruvllm_chat_format` with `messages` (`[{role, content}]`) and `template`: one of `llama3`, `mistral`, `chatml`, `phi`, `gemma`, or a model id to auto-detect. Both fields are required. It returns the formatted string. It needs `@ruvector/ruvllm-wasm` (`npm i -g @ruvector/ruvllm-wasm`); without it the call fails with an install hint.
2. **Create a router** (optional) — `ruvllm_hnsw_create` with `dimensions` (for example 384), `maxPatterns`, `efSearch`. The WASM router holds about 11 patterns at most (ruvllm-wasm v2.0.1), so use it for routing between a few named intents, not for document retrieval. For real document search use `memory_search` or the `ruflo-rag-memory` plugin.
3. **Add patterns** — `ruvllm_hnsw_add` with `routerId`, a `name`, and an `embedding` whose length equals the router's dimensions (you produce the embedding, for example with `embeddings_generate`).
4. **Route a query** — `ruvllm_hnsw_route` with `routerId`, the query `embedding`, and `k` (default 3).
5. **Check status** — `ruvllm_status` shows whether the WASM module loaded. Routers live in the MCP server process only; they are gone when it restarts.
© ruvnet · MIT · ruvnet/ruflo
