Développement
SocratiCode
Enterprise-grade codebase intelligence for AI agents — hybrid semantic search over 40M+ LOC, locally.
★ 3,3 k étoilesAGPL-3.0
Voir le guide →grepai est un projet de développement open source, créé par yoanbernabeu (yoanbernabeu/grepai sur GitHub). Il compte 1 902 étoiles sur GitHub et 161 forks, il est publié sous la licence MIT et a été mis à jour pour la dernière fois sur GitHub le 21 septembre 2026.
grepai: semantic code search and call graphs for AI agents, running 100% locally on your machine.
Decide which embedding provider to use; a local Ollama model keeps everything on your machine.
Install grepai with Homebrew or the install script, and pull the embedding model if you use Ollama.
Initialise grepai in your project and start the watch daemon to build and maintain the index.
Search semantically in plain language and trace callers before changing a function; optionally connect it to your agent as an MCP tool.
Natural-language queries find conceptually related code even when names differ, for example finding session handling when you search for authentication logic.
Trace who calls a function before changing it, reducing breakage from blind edits.
With local embedding providers, your code never leaves your machine, suiting private and regulated codebases.
A file watcher keeps the index up to date automatically as you edit.
Designed to feed coding agents relevant context instead of raw grep output, and exposes itself as an MCP tool agents can call directly.
Generated completions for zsh, bash, fish, and PowerShell cover commands, flags, and dynamic values like workspace and provider names.
Most users install the released binary instead of building from source.
git clone https://github.com/yoanbernabeu/grepai.git
cd grepaiInstall via Homebrew (or the README's install script) and pull the default embedding model.
brew install yoanbernabeu/tap/grepai
ollama pull nomic-embed-textRun these inside your project.
grepai init
grepai watch
grepai search "error handling"It is created and maintained by Yoan Bernabeu.
Yes, grepai is MIT-licensed and free. With Ollama or LM Studio embeddings there is no per-query cost at all.
No paid key is required: Ollama is the recommended default and runs locally. OpenAI embeddings are an optional alternative that would use your key.
ripgrep matches exact text and regex; grepai matches meaning through embeddings and adds call-graph tracing, at the cost of maintaining a local index.
Install it, run ollama pull nomic-embed-text, then grepai init and grepai watch in your project, and search in plain language.
Enterprise-grade codebase intelligence for AI agents — hybrid semantic search over 40M+ LOC, locally.
Fast, accurate semantic code search for agents — about 99% fewer tokens than grep plus read.
Lightweight AST-based semantic code search CLI (ccc) that makes coding agents faster and cheaper
Semantic code search for AI — ripgrep speed with tree-sitter AST parsing.
Eleven AI agents that enforce WCAG 2.2 AA accessibility in your code.
Fast, local, indexed code search built for AI assistants.
Les autres projets de la catégorie Développement avec le plus d'étoiles du catalogue LibreAI — la sélection réaliste d'alternatives si grepai ne vous convient pas.
The open source coding agent
The model-definition framework behind modern AI — thousands of pretrained transformers for text, vision, and audio.
Anthropic's agentic coding tool that lives in your terminal and works across your whole codebase
Cross-platform desktop all-in-one manager for Claude Code, Codex, OpenCode and other coding agents
OpenAI's lightweight coding agent that runs locally in your terminal
Projets de la catégorie Développement proches dans le catalogue, les plus proches en étoiles de grepai — continuez à explorer de part et d'autre.
AI-Powered, Non-Intrusive Terminal Assistant
Instant AI code reviews
Shared memory for coding agents: lessons, repo knowledge, and your preferences, carried across sessions.
A native Rust port of the Pi coding agent: single binary, zero unsafe code, and hardened extensions.
Seamless integration between Claude Code AI assistant and Neovim
Turn any GitHub repo into one formatted text file — perfect context for AI.
Gratuit et ouvert : récupérez le code source, mettez une étoile au dépôt et commencez à créer.