Help us improve
Share bugs, ideas, or general feedback.
Share bugs, ideas, or general feedback.
Share bugs, ideas, or general feedback.
Logic-LLM integration for Claude Code - enables symbolic reasoning with Prolog, Clingo/ASP solvers. Converts documents to logic programs for complex reasoning, legal document processing, and multi-turn inference.
npx claudepluginhub newjerseystyle/plugin-logic-llmSpecialized agent for Answer Set Programming (ASP) with Clingo solver
DMN (Decision Model and Notation) compliance validator for Prolog programs
Agent for converting natural language documents to symbolic logic programs
Specialized agent for First-Order Logic theorem proving with Vampire, E prover, and Prover9
Specialized agent for Prolog-based logical reasoning and knowledge base management
Convert natural language documents into symbolic logic programs.
Batch ingest documents into the symbolic knowledge base with automatic incremental updates.
Query symbolic knowledge bases using natural language or formal logic syntax.
Register and manage predicates, facts, and rules in the logic knowledge registry.
Admin access level
Server config contains admin-level keywords
Modifies files
Hook triggers on file write and edit operations
Share bugs, ideas, or general feedback.
Own this plugin?
Verify ownership to unlock analytics, metadata editing, and a verified badge.
Sign in to claimOwn this plugin?
Verify ownership to unlock analytics, metadata editing, and a verified badge.
Sign in to claimBased on adoption, maintenance, documentation, and repository signals. Not a security audit or endorsement.
First Principles Framework (FPF) for structured reasoning using workflow command pattern. Implements ADI (Abduction-Deduction-Induction) cycle via propose-hypotheses workflow with fpf-agent for hypothesis generation, logical verification, empirical validation, and auditable decision-making. Includes utility commands for status, query, decay, actualize, and reset.
Deep analysis mode with extended reasoning for complex problems
Knowledge base with semantic search, document storage, and automatic summarization. Perfect for domain-specific knowledge management.
Use this agent when you need to create or convert prompts using the SYMBO (symbolic) notation system. This agent MUST be activated whenever generating SYMBO prompts or converting existing prompts to symbolic format. Examples: <example>Context: User wants to create a symbolic prompt for a task management system. user: 'Create a SYMBO prompt for a project task tracker with memory and learning capabilities' assistant: 'I'll use the angelos-symbo agent to create this symbolic prompt following SYMBO notation rules' <commentary>The user is requesting a SYMBO prompt, so the angelos-symbo agent must be used to ensure proper symbolic notation and rule compliance.</commentary></example> <example>Context: User has a natural language prompt they want converted to SYMBO format. user: 'Convert this prompt to SYMBO notation: You are an AI that helps with code reviews by analyzing code quality, suggesting improvements, and tracking common issues across projects' assistant: 'I need to convert this to SYMBO notation using the angelos-symbo agent' <commentary>Since this involves SYMBO prompt generation/conversion, the angelos-symbo agent must be activated.</commentary></example>
Hot-reloadable versioned prompts with easy tools for prompt engineering, chain workflows, quality gates. Symbolic syntax: >>prompt --> >>chain @framework :: 'gate'
Multi-model consensus engine integrating OpenAI Codex CLI, Gemini CLI, and Claude CLI for collaborative code review and problem-solving.
Monte Carlo Tree Search (MCTS-LLM) for intelligent problem solving in research, planning, and coding tasks
Multi-agent code generation with retrieval, planning, coding, and debugging pipeline
Uses power tools
Uses Bash, Write, or Edit tools
Uses power tools
Uses Bash, Write, or Edit tools
Share bugs, ideas, or general feedback.
A Claude Code plugin that integrates symbolic reasoning capabilities using multiple solvers, based on the Logic-LLM methodology.
This plugin enables Claude Code to:
| Solver | MCP Server | Best For | Repository |
|---|---|---|---|
| Prolog | prolog-mcp | Deductive reasoning, Horn clauses | adamrybinski/prolog-mcp |
| Clingo | clingo-mcp | Constraints, optimization, ASP | NewJerseyStyle/clingo-mcp |
| Z3/SMT | z3smt-mcp | Arithmetic, satisfiability, verification | NewJerseyStyle/z3smt-mcp |
| FOL Prover | folprover-mcp | First-order logic (∀, ∃), theorem proving | NewJerseyStyle/folprover-mcp |
| Pyke | pyke-mcp | Forward/backward chaining, Python integration | NewJerseyStyle/pyke-mcp |
| Problem Type | Recommended Solver |
|---|---|
| Deductive chains (ProntoQA, ProofWriter) | Prolog, Pyke |
| Constraints with optimization | Clingo |
| First-order logic with quantifiers (FOLIO) | FOL Prover |
| Constraint satisfaction (LogicalDeduction) | Clingo, Z3 |
| Analytical reasoning (AR-LSAT) | Z3 |
| Arithmetic constraints | Z3 |
| Legal/policy with exceptions | Clingo, Prolog |
The plugin supports generating DMN-compatible Prolog for business rule systems:
/logic-convert --format prolog --dmn-compatible --test ./decision-rules.txt
DMN Constraints Enforced:
Validation via prolog-mcp:
See examples/dmn-credit-approval.md for a complete example.
Important: This plugin extends the original Logic-LLM approach:
| Feature | Original Logic-LLM | This Plugin |
|---|---|---|
| Scope | Q&A only | Q&A + code generation + document conversion |
| Memory | Per-query | Persistent registry across sessions |
| Context | Single query | Multi-turn with symbolic memory |
| Input Size | Within context | Unlimited via chunked processing |
| Code Reuse | None | Registry-based predicate reuse |
| Solvers | Per-dataset | All solvers available, auto-selection |
| Output | Answers | Answers + reusable logic programs |
The plugin stores generated predicates, facts, and rules in a registry that persists across sessions. This enables:
Simple setup is to add in Claude Code CLI
/plugin marketplace add NewJerseyStyle/Claude-plugins-marketplace
/plugin install logic-llm
:warning: This method is not well tested, manual install is more reliable
# Clone the plugin repository
git clone https://github.com/NewJerseyStyle/plugin-logic-llm.git
# Install to user scope (available across all projects)
claude plugin install ./plugin-logic-llm --scope user
# Or install to project scope (shared with team via git)
claude plugin install ./plugin-logic-llm --scope project
cd plugin-logic-llm/servers/registry-server
npm install
npm run build
# Prolog
# npx -y github:adamrybinski/prolog-mcp
# Clingo
npm install clingo-mcp
# Z3/SMT
pip install z3smt-mcp
# FOL Prover
pip install folprover-mcp
# Pyke
pip install git+https://github.com/evertrol/pyke3 pyke-mcp
See CONFIGURATION.md for detailed setup instructions.
/logic-convertConvert documents to symbolic logic programs.