# lychee

> lychee — hermionenoncyclical848-lychee. Use this tool when you need to efficiently manage and orchestrate local Large Language Models (LLMs) for tasks like natural language processing and machine learning. It solves problems related to model complexity, output structuring, and resource allocation by providing features like multi-model pipelines, memory management, and load balancing. Ideal for use cases involving multiple LLMs, it takes in model configurations and input data, and outputs structured results, integrating with git for version control.

Canonical page: https://skillsregistry.net/skills/hermionenoncyclical848-lychee  
JSON: https://api.skillsregistry.net/v1/skills/hermionenoncyclical848-lychee

## Description

Orchestrate local LLMs with Ollama using multi-model pipelines, structured outputs, memory management, and load balancing.

## Trust

- **Trust score (0–1):** 0.50
- **Verification tier:** unverified

## Facts

- **Version:** 1.0.0
- **Skill type:** atomic
- **Execution layer:** container
- **Runtime environment:** vm
- **Updated:** 2026-09-28

## Source

- **Source listing:** [GitHub](https://github.com/hermionenoncyclical848/lychee)

## Use it

Resolve this record through the SkillsRegistry MCP server (no auth, read-only):

```
claude mcp add --transport http --scope user skillsregistry https://api.skillsregistry.net/mcp
```

```json
{
  "jsonrpc": "2.0",
  "id": 1,
  "method": "tools/call",
  "params": {
    "name": "get_skill",
    "arguments": {
      "slug": "hermionenoncyclical848-lychee"
    }
  }
}
```

REST: `GET https://api.skillsregistry.net/v1/skills/hermionenoncyclical848-lychee` · pull for local use: `GET https://api.skillsregistry.net/v1/skills/hermionenoncyclical848-lychee/pull`

---
SkillsRegistry indexes agent skills from public registries and GitHub. Skills we have analysed are scanned with Circle-IR and scored on six dimensions; each listing states its scan coverage. More: https://skillsregistry.net/llms.txt