Open Source Coding Agents
Discover 109 open source Coding Agent repositories from GitHub, each with an analysis of what it does, key features, use cases and alternatives. Coding Agent projects here are most often combined with AI Agents, Developer Tools and CLI. Last updated October 4, 2026.
109 repositories · updated October 4, 2026
magic-context: Manage Coding Agent Context and Memory
Magic Context manages coding-agent context and project memory across sessions, reducing reliance on disruptive compaction and manual note-taking. It is designed for developers using OpenCode, Pi, or Oh My Pi who want persistent, searchable project knowledge.
aft: Give Coding Agents IDE-Style Code Tools
AFT equips OpenCode, Pi, and OMP coding agents with symbol-aware code navigation and edits, semantic search, and persistent shell tasks. It is a fit for developers who want agents to work across a codebase more precisely and with less noisy tool output.
deja-vu: Search Coding Agent Session History
deja-vu indexes existing coding-agent conversations into a shared, local search tool. It helps developers recover past decisions, fixes, and commands across supported agents without requiring an LLM or embedding service.
OmniRoute: Route AI Tools Through One Gateway
OmniRoute is a self-hosted AI gateway that gives coding tools and applications one endpoint for many model providers. It adds configurable routing, quota-aware fallback, and optional token compression, with support for free and paid provider accounts.
free-claude-code: Connect Coding Agents to Multiple AI Models
Free Claude Code is a local proxy and admin interface for routing coding agents to cloud and local language models. It suits developers who want one place to configure providers, switch models, and connect multiple coding clients.
ouroboros: Run a Persistent, Self-Modifying AI Agent
Ouroboros is a Python AI agent for desktop or headless use, designed to retain identity and memory across tasks and restarts. It can work in external projects, coordinate specialist agents, and evolve its own implementation through reviewed changes.
DeepSeek-Reasonix: Run a Local AI Coding Agent
DeepSeek-Reasonix is a Go-based coding agent for complex software engineering tasks, available through a CLI, desktop app, browser, and editor integrations. It combines configurable model providers with tools and controls for managing longer agent runs.
ds4: Run DeepSeek and Other LLMs Locally
ds4, also called DwarfStar, is a native inference engine for selected open-weight language models on Apple Metal, NVIDIA CUDA, and AMD ROCm. It suits users with capable local hardware who want model-specific inference, serving, or coding-agent workflows.
loopx: Govern Long-Running AI Agent Work
LoopX is a local-first control plane that preserves goals, decisions, tasks and evidence across agent sessions. It helps individuals and teams coordinate long-running work while keeping continuation, handoffs and human approval visible.
grok-build: Code with an Interactive Terminal AI Agent
Grok Build is a Rust-based coding agent that works in a full-screen terminal interface or in headless scripts and CI. It can inspect and edit a codebase, run shell commands, search the web, and manage longer tasks.
autoresearch: Run Goal-Driven Improvement Loops with Coding Agents
Autoresearch turns Claude Code, OpenCode, or OpenAI Codex into an agent that makes focused changes, verifies them against a metric, and keeps or reverts each attempt. It suits developers who can define a goal and a mechanical way to measure progress.
agent-memory: Give Coding Agents Persistent, Tiered Memory
agent-memory turns coding-session notes into organized, tiered knowledge and generates an AGENTS.md section with relevant context for new threads. It is a TypeScript tool for teams and developers who want agents to reuse knowledge without relying on manual searches.