FOL Prover MCP Server
An MCP server for first-order logic theorem proving supporting multiple provers like Vampire, E, and Prover9, with built-in simple prover, session management, and TPTP export.
README
FOL Prover MCP Server
An MCP (Model Context Protocol) server for First-Order Logic theorem proving using Vampire, E, and Prover9.
Features
- Multiple Provers: Support for Vampire, E (eprover), Prover9, and built-in simple prover
- Built-in Prover: Simple resolution-based prover requires no external installation
- FOL Parsing: Parse and validate first-order logic formulas with Unicode notation
- Session Management: Build proofs incrementally with named sessions
- TPTP Export: Convert problems to standard TPTP format
- Automatic Fallback: Try multiple provers if one fails
Installation
Prerequisites
The server includes a built-in simple prover that works without any external installation. For more complex proofs, install one of the following theorem provers:
Vampire (recommended):
# Linux (Ubuntu/Debian)
sudo apt-get install vampire
# macOS (with Homebrew)
brew install vampire
# Or download from: https://github.com/vprover/vampire
E Prover:
# Linux (Ubuntu/Debian)
sudo apt-get install eprover
# macOS
brew install eprover
# Or download from: https://wwwlehre.dhbw-stuttgart.de/~sschulz/E/E.html
Prover9:
# Download from: https://www.cs.unm.edu/~mccune/prover9/
Install the MCP Server
pip install folprover-mcp
Or install from source:
git clone https://github.com/folprover-mcp/folprover-mcp
cd folprover-mcp
pip install -e .
Configuration
Add to your MCP client configuration:
Claude Desktop
Add to ~/.config/claude/claude_desktop_config.json (Linux/macOS) or %APPDATA%\Claude\claude_desktop_config.json (Windows):
{
"mcpServers": {
"folprover": {
"command": "folprover-mcp"
}
}
}
VS Code with Continue
Add to your Continue configuration:
{
"mcpServers": {
"folprover": {
"command": "folprover-mcp"
}
}
}
Usage
FOL Notation
The server supports standard FOL notation with Unicode operators:
| Symbol | Meaning | Example |
|---|---|---|
∀ |
Universal quantifier | ∀x P(x) |
∃ |
Existential quantifier | ∃x P(x) |
∧ |
Conjunction (AND) | P(x) ∧ Q(x) |
∨ |
Disjunction (OR) | P(x) ∨ Q(x) |
→ |
Implication | P(x) → Q(x) |
↔ |
Biconditional | P(x) ↔ Q(x) |
¬ |
Negation | ¬P(x) |
⊕ |
Exclusive OR | P(x) ⊕ Q(x) |
You can also use ASCII alternatives:
forallorallfor∀existsfor∃&orandfor∧|ororfor∨->orimpliesfor→<->orifffor↔~ornotfor¬
Tools
prove
Execute a FOL proof directly:
{
"premises": [
"∀x (Human(x) → Mortal(x))",
"Human(socrates)"
],
"conclusion": "Mortal(socrates)",
"prover": "vampire"
}
add_premise
Add a premise to the current session:
{
"premise": "∀x (Human(x) → Mortal(x))"
}
set_conclusion
Set the conclusion to prove:
{
"conclusion": "Mortal(socrates)"
}
prove_session
Prove using the current session's premises and conclusion:
{
"prover": "vampire"
}
parse_formula
Parse and validate a FOL formula:
{
"formula": "∀x (P(x) → Q(x))"
}
convert_to_tptp
Convert a problem to TPTP format:
{
"premises": ["∀x (P(x) → Q(x))", "P(a)"],
"conclusion": "Q(a)"
}
list_provers
List available theorem provers:
{}
Session Management
create_session: Create a new named sessionlist_sessions: List all active sessionsswitch_session: Switch to a different sessionget_session: Get current session stateclear_session: Clear all premises and conclusionremove_premise: Remove a premise by index
Examples
Example 1: Classic Syllogism
Premises:
- All humans are mortal:
∀x (Human(x) → Mortal(x)) - Socrates is human:
Human(socrates)
Conclusion: Socrates is mortal: Mortal(socrates)
Result: Theorem (True)
Example 2: Set Theory
Premises:
- If x is a subset of y and y is a subset of z, then x is a subset of z:
∀x ∀y ∀z ((Subset(x,y) ∧ Subset(y,z)) → Subset(x,z)) - A is a subset of B:
Subset(a, b) - B is a subset of C:
Subset(b, c)
Conclusion: A is a subset of C: Subset(a, c)
Result: Theorem (True)
Example 3: With Counter-model
Premises:
- Some birds can fly:
∃x (Bird(x) ∧ CanFly(x))
Conclusion: All birds can fly: ∀x (Bird(x) → CanFly(x))
Result: Not a theorem (False - there's a counter-model where some bird can't fly)
Architecture
folprover-mcp/
├── src/folprover_mcp/
│ ├── __init__.py
│ ├── server.py # MCP server implementation
│ ├── provers.py # Prover interfaces (Vampire, E, Prover9, Simple)
│ ├── simple_prover.py # Built-in resolution prover
│ ├── fol_parser.py # FOL formula parser
│ └── tptp_converter.py # TPTP format converter
├── tests/ # Test suite
├── examples/ # Example proof problems
├── pyproject.toml
└── README.md
References
- Vampire Theorem Prover
- E Theorem Prover
- Prover9
- TPTP Problem Library
- Logic-LLM - Original inspiration
- MCP Specification
License
MIT License
推荐服务器
Baidu Map
百度地图核心API现已全面兼容MCP协议,是国内首家兼容MCP协议的地图服务商。
Playwright MCP Server
一个模型上下文协议服务器,它使大型语言模型能够通过结构化的可访问性快照与网页进行交互,而无需视觉模型或屏幕截图。
Magic Component Platform (MCP)
一个由人工智能驱动的工具,可以从自然语言描述生成现代化的用户界面组件,并与流行的集成开发环境(IDE)集成,从而简化用户界面开发流程。
Audiense Insights MCP Server
通过模型上下文协议启用与 Audiense Insights 账户的交互,从而促进营销洞察和受众数据的提取和分析,包括人口统计信息、行为和影响者互动。
VeyraX
一个单一的 MCP 工具,连接你所有喜爱的工具:Gmail、日历以及其他 40 多个工具。
graphlit-mcp-server
模型上下文协议 (MCP) 服务器实现了 MCP 客户端与 Graphlit 服务之间的集成。 除了网络爬取之外,还可以将任何内容(从 Slack 到 Gmail 再到播客订阅源)导入到 Graphlit 项目中,然后从 MCP 客户端检索相关内容。
Kagi MCP Server
一个 MCP 服务器,集成了 Kagi 搜索功能和 Claude AI,使 Claude 能够在回答需要最新信息的问题时执行实时网络搜索。
e2b-mcp-server
使用 MCP 通过 e2b 运行代码。
Neon MCP Server
用于与 Neon 管理 API 和数据库交互的 MCP 服务器
Exa MCP Server
模型上下文协议(MCP)服务器允许像 Claude 这样的 AI 助手使用 Exa AI 搜索 API 进行网络搜索。这种设置允许 AI 模型以安全和受控的方式获取实时的网络信息。