Développement
ctx
Hunt down the lowest-cost AI coding setup your repo can rely on.
★ 588 étoilesMIT
Voir le guide →lean-ctx est un projet de développement open source, créé par yvgude (yvgude/lean-ctx sur GitHub). Il compte 3 870 étoiles sur GitHub et 351 forks, il est publié sous la licence Apache-2.0 et a été mis à jour pour la dernière fois sur GitHub le 8 octobre 2026.
LeanCTX — the Rust context gateway that controls what your AI coding tools can see.
Decide the first concrete outcome you want from it, so you can pick the right example, template or mode instead of exploring everything at once.
Clone the repository (or install the published package) and skim its README quick start for the requirements for your platform.
Run the documented install step and set any required configuration — such as provider keys, data sources or local services — before the first run.
Execute the smallest documented example end to end, confirm the output is what you expected, then adapt it to your real data or workflow.
16 read modes (full, map, signatures, diff, density and more), 85+ shell-output compression patterns and cached re-reads that return compact references.
An IntentEngine classifies query complexity and a ModePredictor learns the best read mode per file type, so simple lookups stay cheap.
Session memory, a temporal knowledge graph and a property graph of imports/calls/exports power impact analysis and search ranking.
Configured path permissions, content filters and context budgets applied before delivery, with per-agent budgets and throttling profiles.
Local token/cost tracking, Shadow Mode savings estimates and CPAO (cost per accepted outcome) reporting via lean-ctx savings.
ctx_proof / ctx_verify export an audit trail of sources, policy decisions and pipeline provenance.
Get the source so you can inspect, configure and run it locally.
git clone https://github.com/yvgude/lean-ctx.git
cd lean-ctxAfter installing LeanCTX (see the README install matrix for your platform), configure a supported integration.
lean-ctx setup # see README for the exact install command for your platformEnable Shadow Mode, work normally, then check what LeanCTX saved.
lean-ctx config set shadow.enabled true
lean-ctx savingsIt is maintained by yvgude on GitHub (yvgude/lean-ctx), where the source, issues and releases live.
Yes — it is free and open source under Apache-2.0, which permits use, modification and commercial redistribution, with licence/notice retention.
No LLM API key of its own — LeanCTX prepares context locally; your existing AI tool keeps using its own model and key.
Compared with raw MCP filesystem servers or manual context stuffing, lean-ctx's distinguishing focus is: leanCTX — the Rust context gateway that controls what your AI coding tools can see. Pick it when that focus matches your need; the alternative may suit you better if its broader or different scope is the priority.
Clone https://github.com/yvgude/lean-ctx (or install its published package), follow the README quick start to install and configure it, then run the smallest documented example — the Run section on this page mirrors those steps.
Hunt down the lowest-cost AI coding setup your repo can rely on.
See what your coding agents did — and what each task cost in tokens.
Context optimization for coding agents: sandbox tool output, keep sessions lean, remember more
Context engineering is the new vibe coding - it's the way to actually make AI coding assistants work. Claude Code is the best for this so that's what this re…
All-in-one open-source backend for agentic coding: database, auth, storage, hosting and AI gateway.
MCP graph engine that compounds context across AI coding assistants.
Les autres projets de la catégorie Développement avec le plus d'étoiles du catalogue LibreAI — la sélection réaliste d'alternatives si lean-ctx 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 lean-ctx — continuez à explorer de part et d'autre.
rmcp: the official Rust SDK for the Model Context Protocol — tokio-based servers and clients
The agent harness for complex codebases inside Codex — project memory, planning and verified completion.
zvec-grep (zg): one local CLI unifying ripgrep-style exact search, BM25, and semantic vector search
The official Java SDK for building Model Context Protocol servers and clients.
Claude Octopus orchestrates Claude, Codex, Copilot, Gemini, and Ollama on one task — and surfaces disagreement
Like htop, but for AI coding agents — watch Claude Code and Codex sessions in real time.
Gratuit et ouvert : récupérez le code source, mettez une étoile au dépôt et commencez à créer.