What is the MRV heuristic?

Minimum Remaining Values (MRV) is a 'fail-fast' strategy that selects the variable with the fewest legal options left to process first, helping the algorithm identify contradictions as early as possible.

Can this skill solve optimization problems?

While it excels at finding valid solutions for constraint-heavy problems, pure optimization problems (finding the 'best' among many valid solutions) are often better suited for local search or linear programming.

When should I avoid using this skill?

Avoid using this skill for problems without clear constraints or for very simple problems where a basic brute-force loop is fast enough to execute.

What is the 'propagate-then-search' pattern?

It is an algorithmic approach that eliminates impossible values using known constraints before making guesses, which significantly reduces the computational search space compared to simple brute force.

Is this based on Peter Norvig's Sudoku solver?

Yes, it follows the implementation principles found in Norvig's famous 'Pytudes' regarding constraint propagation and recursive backtracking search.

Constraint Satisfaction Solver

Name: Constraint Satisfaction Solver
Author: jimmc414

byjimmc414

•

データサイエンスとML

Solves complex logic puzzles and scheduling problems using Peter Norvig's propagate-then-search algorithmic pattern.

This skill equips Claude with specialized algorithmic patterns for solving Constraint Satisfaction Problems (CSPs) by implementing a sophisticated propagate-then-search approach. It enables the efficient resolution of complex puzzles and resource allocation tasks such as scheduling, Sudoku, and logic riddles by leveraging the Minimum Remaining Values (MRV) heuristic to fail fast and reduce search space. This skill is ideal for any scenario involving variables, domains, and strict constraints where brute force would be computationally expensive.

主な機能

01Minimum Remaining Values (MRV) heuristic to prioritize constraints

02Automated constraint propagation to eliminate impossibilities early

031 GitHub stars

04Propagate-then-search pattern for efficient state reduction

05Optimized handling for Sudoku, N-queens, and Einstein's riddles

06Recursive depth-first search with state-copying for safe branching

ユースケース

01Generating automated schedules for classes, shifts, or tournament brackets

02Solving high-difficulty Sudoku puzzles or similar grid-based logic games

03Resolving complex resource assignment and placement problems with strict rules

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add jimmc414/claude-code-plugin-marketplace solve-constraint-puzzle

For use in Claude.ai and ChatGPT

Download Skill