lean-mcp

lean-mcp is a skill for Claude Code, Codex from Verified-zkEVM/clean. It costs 0 tokens per session (978 once invoked), scanned A, original, MIT.

A connection between an AI assistant and Lean's language server, which can show the exact proof goal at a chosen line of a Lean file.

In plain words
What is it for?
Use it to inspect proof states, understand tactic behavior, and check proofs without waiting for a complete project build.
Why use it?
It helps you see how tactics such as `rw` and `simp` change a proof when the result is not what you expect.

Skill for Claude CodeCodex

Written for no agent in particular: nothing here depends on one.

Good fit Use it to inspect proof states, understand tactic behavior, and check proofs without waiting for a complete project build.

Compare 6 skills from other repositories ↓
Install with agentmods
npx agentmods add skills/verified-zkevm/clean/lean-mcp
Install

Getting it into your agent

One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.

Any agent
npx skills add Verified-zkEVM/clean --skill lean-mcp
Clone the repo
git clone --depth 1 https://github.com/Verified-zkEVM/clean

Made for: Claude Code, Codex.

Wrote this? Show the measurements

A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.

agentmods badge for lean-mcp

README.md
[![agentmods](https://agentmods.dev/badge/skills/verified-zkevm/clean/lean-mcp/github.svg)](https://agentmods.dev/skills/verified-zkevm/clean/lean-mcp)
Your own site
<a href="https://agentmods.dev/skills/verified-zkevm/clean/lean-mcp"><img src="https://agentmods.dev/badge/skills/verified-zkevm/clean/lean-mcp/github.svg" alt="Measured on agentmods" height="20"></a>

Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.

agentmods 80×15 button for lean-mcp

Your own site · 80×15
<a href="https://agentmods.dev/skills/verified-zkevm/clean/lean-mcp"><img src="https://agentmods.dev/badge/skills/verified-zkevm/clean/lean-mcp.svg" alt="Reviewed on agentmods" width="80" height="20"></a>
Per session 0 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 978 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. A grade says what 26 rules found in the file — not that it is safe. Third-party audits
  • NVIDIA SkillSpector pass 7 Sept 2026
How audits are shown
Origin original No closer match found in the catalogue.
Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

ModelPer sessionOnce invoked
Fable 5.1 $0.00000 $0.00978
Opus 5.5 $0.00000 $0.00391
Sonnet 5.5 $0.00000 $0.00196
Haiku 4.5 $0.00000 $0.00098

Measured 2mo ago against content hash 526291464e99, method: parsed. Prices are Anthropic first-party input rates as of 2026-10-07, from the pricing page.

Security

Grade A, and why

lean-mcp scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 2mo ago.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

skills/lean-mcp/SKILL.md · 119 lines

How it starts

The opening of the file, as written. The whole thing — 119 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Lean LSP via mcporter - Skill

This skill enables interactive inspection of Lean proof states using the Lean Language Server Protocol via mcporter.

When to use this skill

  • You're stuck on a Lean proof and need to see the exact goal state after simplifications
  • Your rw or simp commands aren't matching the expected pattern
  • You want to understand what a tactic is actually doing
  • You need to verify a proof compiles without waiting for full lake build

Setup (one-time)

npm install -g mcporter
# uv should already be installed (check with: which uv)

The config/mcporter.json in the repo is already configured for the lean-lsp MCP server.

Key Commands

1. View proof state at a line (MOST IMPORTANT)

mcporter call lean-lsp.lean_goal \
  file_path:/absolute/path/to/File.lean \
  line:123

Returns:

  • goals_before: Goal state at the start of the line
  • goals_after: Goal state at the end of the line (after tactic executes)
  • Shows exactly how the tactic transforms the goal

Example:

cd /mnt/data-2tb/tmp/clean
mcporter call lean-lsp.lean_goal \
  file_path:/mnt/data-2tb/tmp/clean/Clean/Examples/FibonacciWithChannels.lean \
  line:728

2. Get diagnostics (errors/warnings)

mcporter call lean-lsp.lean_diagnostic_messages \
  file_path:/absolute/path/to/File.lean \
  start_line:100 end_line:150

3. Try multiple tactics without modifying the file

mcporter call lean-lsp.lean_multi_attempt \
  file_path:/path/to/File.lean \
  line:123 \
  'snippets:["simp", "rw [foo]", "exact bar"]'

Returns the goal state after each tactic, so you can see which one makes progress.

4. Get type/documentation for a symbol

mcporter call lean-lsp.lean_hover_info \
  file_path:/path/to/File.lean \
  line:123 column:10

Column should be at the START of the identifier you want to inspect.

5. Search for theorems/lemmas

# Natural language search via LeanSearch
mcporter call lean-lsp.lean_leansearch \
  query:"sum of two even numbers is even" \
  num_results:5

# Type signature search via Loogle  
mcporter call lean-lsp.lean_loogle \
  query:"(?a → ?b) → List ?a → List ?b" \
  num_results:8

# Semantic search via LeanFinder
mcporter call lean-lsp.lean_leanfinder \
  query:"I have h : n < m and need n + 1 < m + 1" \
  num_results:5

Read the full file on GitHub · 119 lines

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

  1. 2mo ago First seen · 119 lines · 0 tokens per session scan A 526291464e99

Subscribe to this mod's changes

lean-mcp is a skill published in the GitHub repository Verified-zkEVM/clean (208 stars, last pushed yesterday), licensed MIT. It costs nothing until one of its globs matches a file; then it loads 978 tokens. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.

Related

lean-mcp compared with skills from other repositories

lean-mcp next to 6 skills from other repositories
ModTypeStarsLast pushLicenceScanTokensWorks with
lean-mcpVerified-zkEVM/cleanSkill208yesterdayMITA0Claude Code, Codex
imagegenopenai/codexSkill133ktodayApache-2.0A113Codex
playwright-climicrosoft/playwrightSkill99kyesterdayApache-2.0A19Claude Code
local-ai-agentsmicrosoft/ai-agents-for-beginnersSkill79k4d agoMITA200Claude Code, Codex
systematic-debuggingobra/superpowersSkill313k3d agoMITA21Claude Code
next-cache-components-adoptionvercel/next.jsSkill143ktodayMITA95Claude Code
chat-pet-sprite-creationmicrosoft/vscodeSkill195ktodayMITA53Claude Code, Codex