ai-code-generation

Открытый бенчмарк LLM: какая нейросеть лучше пишет код 1С:Предприятие (BSL). Объективная оценка LLM по методике SMOP с реальным исполнением в 1С — Claude, GPT, Gemini, DeepSeek, YandexGPT, GigaChat.

AI Code Stitcher - Paste AI Generated Code - It Just Works! Copy output from Claude, Cursor, ChatGPT, Aider or any other LLM and AI Code Stitcher intelligently applies it to the correct locations in your existing codebase.

AI-powered Python CLI tool that automates the entire software development lifecycle using the Claude Code SDK. From specification to deployment, Claude Code Builder

Advanced Claude Code Tutorial

Sublime Text Plugin which connects to a local Model Context Protocol (MCP) server for AI-powered code generation, review, refactoring, and translation using OpenAI or Gemini.

OutcomeOps: Context Engineering for AI-Assisted Development

This repository allows the replication of our study "Human-Written vs. AI-Generated Code: A Large-Scale Study of Defects, Vulnerabilities, and Complexity" accepted for publication at The 36th IEEE International Symposium on Software Reliability Engineering (ISSRE 2025).

Codehive Genie is an AI-powered coding assistant integrated into the Codehive platform, designed to simplify coding tasks such as code generation, bug fixing, and optimization. Powered by Llama 3.1 and deployed on Vercel, it supports multiple programming languages, including Python, C, Java, JavaScript, TypeScript, and C++.

Code-validation kernel for AI-assisted development: source-bound signed evidence, deterministic acceptance gates and a verifiable audit trail. MIT.

AI code generation framework with 12-agent review system, automated testing (pytest/jest/go), and iterative auto-fix. Natural language → production code with security scanning, Docker isolation, and multi-platform deploy configs.

Codehive Genie is an AI-powered coding assistant integrated into the Codehive platform, designed to simplify coding tasks such as code generation, bug fixing, and optimization. Powered by Llama 3.1 and deployed on Vercel, it supports multiple programming languages, including Python, C, Java, JavaScript, TypeScript, and C++.

Датасет бенчмарка PRISM: 1225 генераций кода 1С (BSL) от 35 нейросети с автооценкой по метрике SMOP. Зеркало: hf.co/datasets/genlab-1c/prism-smop

Sample edition of The Stack Enriched: annotated, secure, and optimized code dataset, this is a sample version

Protocol for building projects that AI models recommend to users — .ai-starter.json spec, registry, and gap scanner.

Adversarial Co-Generation Engine — concurrent Builder+Breaker agents for production-grade AI code security

Open code golf solutions archive with verified answers, AI/human submissions, and reproducible code.golf runner automation.

A payment system built by an AI agent from a frozen spec, zero hand written tests, guarded by a four layer verification pyramid (structural, behavioral, agent judgment, mutation). 73.1 percent mutation kill rate.

Spec-first verification loop for AI-generated code: behavioral E2E tests as an executable specification.

Certus is a finetuned model that generates machine-checkable correctness certificates for Python functions, plus a checker that verifies them with counterexample search. 83% of certificates survive Hypothesis testing on unseen code.

Adversarial Prover/Skeptic code-generation loop, verified by real sandboxed execution — not an LLM's opinion