# RoCQ

> Use this tool when you need to integrate advanced logical reasoning capabilities into your workflow, particularly for formal methods, theorem proving, and verified programming. RoCQ solves problems related to automated dependent type checking, inductive type definition, and property proving, accepting natural language inputs and providing detailed feedback on type errors and failed proofs. It is ideal for researchers and developers who require rigorous mathematical reasoning and software correctness proofs without directly interacting with complex formal verification syntax.

Canonical page: https://skillsregistry.net/skills/tyler-blaine-hall-rocq  
JSON: https://api.skillsregistry.net/v1/skills/tyler-blaine-hall-rocq

## Description

This MCP server, developed by Tyler Blaine Hall, integrates the Coq proof assistant to provide advanced logical reasoning capabilities. Built with Python and leveraging the MCP SDK, it offers tools for automated dependent type checking, inductive type definition, and property proving using custom tactics and automation. The implementation focuses on bridging natural language inputs with Coq's formal verification system through an XML protocol, enabling detailed feedback on type errors and failed proofs. It's particularly useful for researchers and developers working on formal methods, theorem proving, and verified programming, allowing for rigorous mathematical reasoning and software correctness proofs without directly interacting with Coq's complex syntax.

## Trust

- **Trust score (0–1):** 0.77
- **Verification tier:** verified
- **Last scanned:** 2026-09-19

## Facts

- **Version:** 1.0.0
- **Skill type:** atomic
- **Execution layer:** mcp-remote
- **Runtime environment:** api
- **Category:** search
- **Updated:** 2026-09-19

## Source

- **Source listing:** [PulseMCP](https://www.pulsemcp.com/servers/tyler-blaine-hall-rocq)
- **Repository:** <https://github.com/angrysky56/mcp-rocq>

## 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": "tyler-blaine-hall-rocq"
    }
  }
}
```

REST: `GET https://api.skillsregistry.net/v1/skills/tyler-blaine-hall-rocq` · pull for local use: `GET https://api.skillsregistry.net/v1/skills/tyler-blaine-hall-rocq/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