# Think Tool

> Use this tool when you need to record, retrieve, and analyze the reasoning processes of Large Language Models during complex problem-solving tasks, such as structured reasoning or step-by-step analysis. It provides a scratchpad for thoughts, retrieval of previous thoughts, and generation of statistics, making it ideal for applications requiring persistent thinking context. The Think Tool is particularly useful for improving problem-solving capabilities in areas like policy verification.

Canonical page: https://skillsregistry.net/skills/iamwavecut-think  
JSON: https://api.skillsregistry.net/v1/skills/iamwavecut-think

## Description

MCP-Think is a Go-based server that implements a 'Think Tool' for Large Language Models, allowing them to record, retrieve, and analyze their reasoning processes during complex problem-solving tasks. Developed by Valeriy Selitskiy, this implementation provides four key tools: recording thoughts as a scratchpad, retrieving previously recorded thoughts, clearing the thought history, and generating statistics about recorded thoughts. The server uses stdio transport for compatibility with various LLM clients including Claude Desktop via Smithery and Cursor, making it particularly valuable for applications requiring structured reasoning, step-by-step analysis, or policy verification where maintaining a persistent thinking context improves problem-solving capabilities.

## Trust

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

## Facts

- **Version:** 1.0.0
- **Skill type:** atomic
- **Execution layer:** mcp-remote
- **Runtime environment:** api
- **Category:** ai-ml
- **Updated:** 2026-04-25

## Source

- **Source listing:** [PulseMCP](https://www.pulsemcp.com/servers/iamwavecut-think)
- **Repository:** <https://github.com/iamwavecut/mcp-think>

## 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": "iamwavecut-think"
    }
  }
}
```

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