{"data":{"kind":"file","path":"README.md","version_id":"j2fxhztldj1y6vbcv8y3roa1","entry":{"name":"README.md","path":"README.md","is_directory":false,"size":1102,"modified_at":"2026-08-05T11:00:50.004000","content_hash":"58d1e7bfad8c3ce9d169e0ea596419d00468caec1dcd41ce0f3721593334eecd"},"entries":[],"content":"# LeetCode Solver\n\nAn environment for evaluating LLM problem-solving ability on classic LeetCode and Data Structures & Algorithms (DSA) problems.\n\n## Overview\n\nThe model receives algorithmic problems (arrays, strings, trees, graphs, dynamic programming, etc.) and must produce correct, efficient Python solutions. Each solution is evaluated by:\n\n1. **Correctness** — running against hidden test cases via subprocess execution\n2. **Code Quality** — checking for proper function signatures, type hints, and docstrings\n3. **Efficiency** — penalizing brute-force approaches when optimal solutions exist\n\n## Problem Categories\n\n- Arrays & Hashing\n- Two Pointers & Sliding Window\n- Stack & Queue\n- Binary Search\n- Linked Lists\n- Trees & Graphs\n- Dynamic Programming\n- Greedy\n- Backtracking\n- Bit Manipulation\n\n## Scoring\n\n- **Correctness (0.6)**: Pass rate on executable test cases\n- **Code Quality (0.2)**: Structure, naming, type hints\n- **Completeness (0.2)**: Edge case handling, input validation\n\n## Usage\n\n```python\nimport verifiers as vf\nenv = vf.load_environment(\"arasaki/leetcode-solver\")\n```\n","encoding":"utf-8","truncated":false,"total_bytes":1102},"status":null}