Agents
553 coding agent harnesses — systems that let an LLM write, modify, debug, or run code. Filter and sort below.
Fully open-source AI coding agent with both terminal UI and beta desktop app; built-in switchable agents (build/plan) via Tab key - the plan agent is read-only,...
An 'agent-managed museum exhibit' developed and maintained autonomously by agent harnesses (LazyCodex/Gajae-Code) with no human intervention; serves as a living...
Anthropic's official terminal-native agentic coding assistant that understands your codebase and operates through natural language; works across terminal, IDE...
OpenAI's official coding agent that runs in the terminal with native sandboxing; can also integrate into VS Code, Cursor, and Windsurf. Supports MCP, plugins,...
Google's official open-source AI agent that brings Gemini models directly into the terminal; free generous tier with Google account; 1M token context window;...
Self-extensible interactive coding agent CLI providing a unified multi-provider LLM API, agent runtime with tool calling and state management, and a terminal UI...
A batteries-included 'super agent harness' orchestrating sub-agents, long-term memory, sandbox-aware execution, and extensible skills; ships with built-in...
Simulates an entire software company as a multi-agent system with specialized roles (product manager, architect, project manager, engineer) following...
Fork of OpenAI's Codex optimized for low-cost models. Features harness emulation (switch between native, claude-code, kimi-code, qwen-code, deepseek-tui,...
Autonomous coding agent available as SDK, IDE extension, or CLI. Features Plan/Act modes, multi-agent teams with coordinator delegating to specialist agents,...
Agentic development environment born from the terminal. Combines a modern terminal with AI-powered coding agent capabilities and support for external CLI agents...
Self-described as 'The OG code generation experimentation platform'; an early tool for specifying software in natural language and having AI write/execute code,...
Open source, extensible AI agent that runs locally on macOS, Linux, and Windows. General-purpose agent for code, research, writing, automation, and data...
AI pair programming in the terminal. Features codebase mapping, git integration with auto-commits, automated linting/testing, voice-to-code, IDE integration via...
Pioneering open-source coding agent available as CLI, VS Code extension, and JetBrains plugin. The repository is no longer actively maintained and is read-only;...
DeepSeek-native AI coding agent engineered around prefix-cache stability for long-running sessions. Distributed as a self-contained static Go binary. Features...
A zero-code multi-agent orchestration platform with visual workflow canvas and drag-and-drop agent orchestration; evolved from a 'Virtual Software Company' into...
Builds fully working, production-ready apps step-by-step using 11 specialized AI agents (Product Owner, Specification Writer, Architect, Tech Lead, Developer,...
AI-powered IDE with multi-agent collaboration, fleets of parallel agents working for hours or days, built-in marketplace with plugins, Plan Mission Control, and...
Coding agent with the IDE wired in. Features 31 built-in tools including LSP-integrated edits, real debugger driving (lldb/dlv/debugpy), persistent Python/JS...
Hybrid architecture code review tool combining deterministic engineering pipelines with an LLM agent, achieving higher precision than general-purpose agents...
Glamourous terminal-based coding agent by Charmbracelet. Supports multi-model LLMs, session-based contexts, LSP integration, MCP servers (stdio/http/sse with...
Feature-rich agentic coding assistant with auto-memory, auto-skills, subagents, agent teams, MCP support, plan mode, hooks, LSP integration, Git worktrees, and...
All-in-one agentic engineering platform for VS Code, JetBrains, or CLI. Features code generation across multiple files, inline autocomplete, self-checking agent...
SpaceXAI's terminal-based AI coding agent running as a full-screen TUI. Understands codebases, edits files, executes shell commands, searches the web, and...
An 'AI-Powered Dev Team, Right in Your Editor' VS Code extension with multiple specialized AI agent modes (Code, Architect, Ask, Debug, custom) and MCP server...
Local, open-source AI app builder for power users. Alternative to v0/Lovable/Replit/Bolt. Cross-platform, runs on your machine, bring your own API keys without...
Uses a Recursive Language Model (RLM) that treats context as variables and tools as function calls in a persistent IPython REPL, combined with a Continual...
Takes a GitHub issue and tries to automatically fix it using your LM of choice. State-of-the-art on SWE-bench among open-source projects (NeurIPS 2024)....
Extreme RAM efficiency (6-14x less than competitors), sub-15ms first frame, human-like agent memory system with semantic vector retrieval, multi-agent swarm...
First open-source implementation of an agentic software engineer; open-source alternative to Devin (Cognition AI). README banner now directs users to its...
Long-lived memory is the organizing primitive: email, meetings, Slack, and assistant conversations are indexed into an Obsidian-style backlinked graph, and...
Paper2Code workflow reproduces research papers into runnable code (75.9% on PaperBench, beating commercial agents by 26 points); Loop Engineering with durable...
2M token effective context window with tree-sitter project maps (30+ languages); cumulative diff sandbox separating AI changes from project files until ready;...
100% open-source, local-first desktop multi-agent cowork app built on CAMEL-AI's framework. Combines single-agent harness with multi-agent workforce...
Persistent cross-session memory via SQLite FTS5, subagent orchestration with parallel execution, goal-driven autonomous loops with independent judge models,...
AI-powered PR review agent where each tool uses a single LLM call for low cost (~30s). Platform-agnostic across 5 git providers (GitHub, GitLab, Bitbucket,...
Completely free with no API key needed; multi-agent orchestration instead of single-model approach; five distinct products (CLI, Desktop, Web, Cloud, Chat);...
Human-centric AI scaffolding agent that generates entire codebases from a product spec. Whole-program coherence via an intermediate shared_dependencies.md step...
Transparent, modular, research-friendly architecture designed for studying AI agent architectures and ablation studies; includes trajectory recording, Lakeview...
AI coding agent platform focused on enabling agents to tackle difficult problems in complex codebases with human-in-the-loop oversight. Repository is explicitly...
Dual-mode shell/agent (toggle with Ctrl-X without leaving CLI), ACP (Agent Client Protocol) support for IDE integration (Zed, JetBrains), VS Code extension, Zsh...
Self-improving architecture — Claude can autonomously identify capability gaps, design, and implement new tools during conversations, making the framework more...
Terminal-native synchronous AI agent powered by the same agentic harness as GitHub's Copilot coding agent; deep GitHub workflow integration (repos, issues, PRs...
AI-driven end-to-end security workflow CLI/SDK: discovers, validates, and auto-patches vulnerabilities, verifies fixes, and opens GitHub PRs. Deep multi-agent...
Open-source replica of the internal coding-agent architecture used by Stripe (Minions), Ramp (Inspect), and Coinbase (Cloudbot); built on LangGraph + Deep...
Clean-room Rust reimplementation of Claude Code as a terminal coding agent (TUI pair programmer). Multi-provider routing with no telemetry/tracking. Agent...
AI coding assistant powered by OpenCode
Originally a GitHub-app AI coding assistant; pivoted to an AI coding assistant for JetBrains IDEs, distributed as a JetBrains plugin. Dual-licensed: MIT for the...
ZSH plugin with colon prefix system (use AI without leaving shell), three-mode architecture (TUI/CLI one-shot/ZSH plugin), semantic code search, conversation...
Automated full-codebase migration across languages/frameworks using LLMs. Spins up a Docker environment for the target language, recursively rebuilds code from...
Agentic IDE for open-source models — browse and build in same tool without context switching; coding agent with full access to console and debugger;...
Local-first AI agent workspace with a dual interface — desktop GUI and terminal TUI share a single local 'kun serve' runtime (not two separate sessions)....
Local-first / offline capable (built-in llama.cpp, Ollama, on-device ONNX MiniLM embeddings for semantic search), no cloud required; per-hunk inline review...
Code-as-action browser agent: instead of predicting one browser action at a time, the LLM writes and debugs Playwright scripts via a terminal, treating the...
Combines LLM with DevOps tools to automate the full software development lifecycle — clarifies natural-language requirements, generates interface docs, writes...
Open-source personal AI assistant shipped as a single binary with no runtime dependencies (runs natively on Linux/macOS/Windows, including native Windows...
Every chat message is automatically a git commit — enabling revert, branch, reset, and squash of AI conversations; watches changes in real-time like a pair...
Open-source agentic platform by Cloudflare for building full-stack apps entirely on the Cloudflare stack (Durable Objects, Dynamic Workers, Artifacts, AI...
Autonomous coding agent embedded directly in VS Code that adapts to any skill level — from beginners to complex system architecture; converts mockups/wireframes...
Mistral's official open-source CLI agent with subagents, hooks (pre_tool/post_tool/post_agent), Skills system following Agent Skills specification, voice mode,...
Enterprise-grade AI development platform (by Chaitin) with cloud development environments, requirement/SPEC management, automated PR/MR code review, and team...
AI app builder by Convex (fork of bolt.diy) that generates full-stack web apps with real backend — built-in database, zero-config auth, file uploads, real-time...
Same project as unit-mesh/auto-dev (GitHub redirect). AI-native multi-agent development platform built on Kotlin Multiplatform (JVM, Android, iOS, JS, WASM)....
AI-native multi-agent development platform built on Kotlin Multiplatform (JVM, Android, iOS, JS, WASM from a single codebase). Core agents (Document, Coding,...
One of the first agent CLIs (Spring 2023), still actively developed. Provider-agnostic and local-first — runs anywhere a terminal runs (SSH, tmux, headless...
Agentic IDE and CLI by AWS/Amazon using spec-driven development — turns prompts into structured Specs then implementation plans. Features 'Powers' (specialized...
Open-source agent operating system (by Tsinghua THUNLP/ModelBest/OpenBMB) centered on 'WorkSpace' isolation — each project gets isolated files, memory, and...
Meta neuro-symbolic AI model called taste-1 that continuously learns your coding preferences and style from interactions; learned taste profile is portable and...
AI coding assistant by Sourcegraph that uses deep codebase context via semantic search for chat, autocomplete, and inline edits; this repo is a public snapshot...
Agentic IDE that builds itself — every surface (desktop app, web app, CLI, HTTP API) is first-class, with live steerable threads, agent hand-offs, and a...
Deeply integrated with Grok API including X (Twitter) live search and web search, built-in computer sub-agent for host desktop automation (macOS), Telegram...
Community-driven open-source coding agent that beat AutoCodeRover on SWE-Bench Lite; offers both terminal UI and Electron-based GUI; supports local models via...
Stateful agent harness where agents self-improve by rewriting their own memory, skills, prompts, and harness; supports subagents, hooks, crons, and messaging...
Terminal-first AI coding agent with a Think → Act → Verify loop; cost-aware intelligent routing (complexity-scored model selection with automatic failover); 50+...
Microsoft's modular AI development assistant. The CLI is one interface on top of a modular platform — bundles ship focused agents you invoke by name, and the...
Open-source AI-native terminal for cloud and infrastructure management that uses natural language to deploy services, troubleshoot, and automate operations...
Lightweight cross-platform code editor built with Tauri (Rust + React) featuring Git support, AI agents, LSP, vim keybindings, integrated terminal, database...
A coding agent harness written in Zig that behaves like a Unix shell rather than a terminal IDE: a ~6 MB binary with microsecond cold start, a shell-like UI...
Small, readable terminal coding agent — a Python port of Pi's minimalist coding agent — that reads, edits, and runs commands with durable JSONL sessions; also...
Coding agent tuned for small local models, built on pi, with plan mode, dispatch sub-coders, per-phase model selection, read-before-edit enforcement, and a...
Open-source terminal agentic coding assistant; bring-your-own-model, no telemetry, skills system (commands/subagents/tools/event triggers), checkpointing, and...
Fully autonomous AI agent running in a sandboxed Docker environment with a built-in browser, text editor, automatic Docker-image picker, and PostgreSQL-backed...
Combines a high-performance Rust agent runtime with a polished desktop app featuring Agentic Mini Apps (each task gets its own dedicated UI bound to live...
A coding assistant using native function calling with DeepSeek-R1 (DeepSeek-Reasoner), featuring visible Chain-of-Thought reasoning, automatic file operations,...
Terminal coding agent with multi-model orchestration that automatically delegates tasks to different models by role (orchestrator, planner, builder, reviewer,...
A trainable multi-agent framework with four specialized modules (Planner, Executor, Verifier, Generator) coordinated through in-the-flow online optimization via...
Desktop and browser application for isolated, parallel agentic development with a custom agent loop inspired by Claude Code. Features a central view on git...
AWS's agentic terminal chat for building applications with natural language; succeeded by the closed-source Kiro CLI. README notes it is no longer actively...
A self-contained Rust coding agent shipped as a single ~15MB dependency-free binary with no model lock-in — 17 provider presets plus a custom catalog layer for...
An automated AI development team (Product Manager, Developer, DevOps) that generates, tests, builds, and deploys microservices from natural language...
A whole dev team of AI agents in your editor (Roo fork)
Community-driven continuation of Roo Code after the original Roo team moved to Roomote. AI-powered VS Code extension providing a team of AI agents in-editor,...
Independent executor-verifier orchestration: 'The agent that wrote the code shouldn't be the one that says it works.' Verifiers don't share the executor's...
100% local-first coding agent bundled with llama.cpp in Docker; zero per-token cost, zero data egress. Auto-detects hardware (NVIDIA/CPU/Apple Silicon),...
'The nanoGPT of coding agents' - radical minimalism at ~1,081 lines across 8 core files, readable in an afternoon. A genuinely runnable teaching tool (86 tests...
Research project (ICML 2024) proposing executable Python code as a unified action space for LLM agents instead of JSON/text. Agents can revise prior actions or...
Minimal high-performance Rust coding agent (~30k LoC, 26MB binary, ~16MB RAM, near-zero idle CPU) - dramatically smaller than JS-based agents like opencode...
Fully local, user-owned terminal coding agent emphasizing control and privacy. Model-agnostic (25+ providers), explicit permission-gated safety model (file...
Development gruntwork is codified as six predefined 'patchflows' (AutoFix, ResolveIssue, PRReview, GenerateDocstring, GenerateREADME, DependencyUpgrade) built...
Coding agent for generating code, fixing bugs, reviewing code, and adding tests, with both interactive and headless modes and a VSCode extension.
Open-source full-stack desktop AI agent combining coding, file operations, terminal commands, tool calls, research, content creation, and multi-agent...
Method-native AI work system for running complex work from intent to verified delivery. Current focus is 'loushang code', a CLI and terminal workbench for...
Open-source agentic telemetry system that ingests traces, logs, and metrics, groups noisy signals into incidents, and uses AI agents to self-heal software....
Open-source agentic orchestration layer emphasizing Transparency (see every token, context file, proposed change), Control (tool approval gates, fork/duplicate...
First bot to autonomously generate pull requests in response to GitHub issues (triggered via label). Built with Guardrails for structured LLM output. Worked...
Multi-step 'Superpowers' workflows that chain different model tiers together for complex tasks; smart model routing automatically selects the fastest model per...
AI backend builder that generates 100% compilable NestJS+Prisma backends via a waterfall pipeline of 40+ specialized agents (Analyze, Database, Interface, Test,...
Multi-Agent collaborative scheduling (codefuse-muAgent framework); repository-level code analysis & generation; document RAG enhanced with knowledge graphs;...
AI coding agent that builds a live dependency graph (genome) of the repo using tree-sitter, then edits code through AST symbol-level operations rather than...
Note: the repo at TrafficGuard/nous contains a project named 'TypedAI' (the repo may have been renamed/redirected). TypeScript-first AI platform for developing...
Intentionally minimal and readable coding agent harness designed for educational purposes to explain the six core components of a coding agent (live repo...
Multi-agent collaborative reasoning across files and commits using a Neo4j-powered Unified Codebase Knowledge Graph and long-term memory (Athena). Full...
Intentionally compact implementation of a Claude Code-like agent loop suitable for study, modification, and as a reference architecture. Available in 5...
Built specifically for DeepSeek models with three flavors (native macOS app, TypeScript/Bun CLI, Rust CLI+GUI) that share configuration, sessions, and tools via...
Terminal agentic coding assistant with team mode (multi-agent parallel collaboration), vulnerability hunting mode, Claude Code Skills compatibility, StatusLine...
Open-source framework inspired by DeepMind's AlphaEvolve that combines evolutionary algorithms with LLM-driven code generation using diff-based mutations and...
A lightweight Rust TUI coding agent built for context-token efficiency — index skeletons, code_execution sandbox, tool_search and model-tier subagents keep cost...
Ships open-sourced, custom-trained autofix models (diff-apply and fix-json on Hugging Face) that repair mangled JSON tool calls and bad diffs from any coding...
Autonomous engineering team runtime where one API call spins up a full team (PM, architect, coders, reviewers, testers) that plans, builds, adapts, and ships...
Editor-agnostic AI pair programming tool using an LSP-like protocol; a central server handles tool call management, multi-LLM interaction, telemetry, and...
4-layer token compression always on (Caveman Mode, Tool Budgets, Read Dedup, RTK); ~2x fewer tokens than Codex CLI on identical tasks; Claude Code-compatible...
AI-native multi-agent website builder — full pipeline from requirement understanding to page planning, design, code generation, CMS content management, SEO,...
Production-oriented Python rebuild of Claude Code (~310K lines); Terminal-Bench 2.1 score 80.9%; 30 model providers vs Claude Code's Claude-only limitation;...
"Open-source generalist agent harness ranked
Local-first agentic IDE for long-running engineering work with durable tasks that persist across restarts; zero-trust philosophy (model output is input, not...
Self-hosted, single-user, mobile-friendly AI workspace packaging the Pi coding agent with a trilingual (English/Chinese/Japanese) streaming web UI,...
Rust terminal coding agent combining LLM-native code understanding, OS-native sandboxing, multi-provider support (including local inference), open protocols...
Compound AI system: multiple models collaborate, each optimized for its role across 5 workflow slots (Normal, Thinking, Compact, Critique, VLM), each...
Safety-first code-agent CLI: spoof-resistant approval previews (Unicode neutralization), folder-trust boundary (fail-closed), deterministic offline command...
Open-source, native GitHub Copilot alternative for VS Code with 100% native VS Code experience (no webviews), BYOK with top AI providers, and flexibility to mix...
Fully offline, free AI coding environment for Mac (also Intel Mac/Linux) using Ollama + local LLM. Built for workshops, students, and beginners. Auto model...
Verifiable coding agent runtime — build engineering processes as explicit, checkable execution graphs (workflows) with stages, checks, artifacts, and approval...
A minimalist, terminal-native coding agent written in C — single lightweight binary with small memory footprint, local models as first-class citizens, respects...
Mathematical coding agent focused on formal theorem proving in Lean. Atomic Lean tools (LeanGoal, LeanCheck, LeanSearch, LeanVerify) for interactive proof...
Intentionally minimal and hackable coding CLI — the antithesis of feature-rich CLIs. Designed as a blueprint/building block that developers modify directly (no...
Autonomous AI agent framework that runs unattended on GitHub Actions — ships features, deploys apps, finds/discloses vulnerabilities, runs research, and writes...
Open-source AI agent for Unity projects — writes C# code, reads/modifies Unity objects & assets. Runs as a standalone Rust/Tauri/Vue.js process (not inside...
Open source AI-powered code editor — a fork of VS Code with AI integration. Submodule is a fork of Continue. Stores code context locally. Familiar VS Code feel...
Runs 100% locally with no OpenAI API key, translating natural language into shell commands via small local models or Ollama, and executes nothing without...
Self-coding agent designed to evolve into a fully autonomous, self-programming AI system. Combines LLMs (GPT-4) with a vector database to document, search, and...
TypeScript library and CLI for building ReAct-style AI agent loops. Three-layer architecture (Foundation, Agent Loop, Coding Agent) keeping provider...
Configuration-driven (YAML) agent harness ('OS for AI agents'); batteries-included (sessions, memory, tools, observability); instant mid-conversation model...
Multi-model AI coding assistant with integrated cross-platform terminal (xterm.js), file attachments (PDF/Word/Excel/OCR), WebSocket real-time updates,...
AI coding agent built on opencode with a dual-speed self-evolution system based on the Pioneer Agent research paper: a fast loop (/evolve) that learns tactics...
Terminal-based pair programming agent with headless mode for scripting/automation, HTTP server mode with OpenAI-compatible API and SSE streaming, shell...
Visual GUI workbench with Miller-column (Finder-style) navigation; sessions are editable, branching trees (not linear transcripts); persistent and stateful...
An open-source Claude Code alternative with distinctive GitHub-native automation: agents trigger from issue comments via /mycoder, run with githubMode for...
Code instruction system where you provide a GitHub issue and it automatically generates an AI-coded Pull Request
Terminal-based AI coding assistant powered by Gemini that uses the model's native function calling to automatically execute file, system, and test operations...
Full Python reimplementation of Claude Code's npm agent architecture with zero external dependencies (pure stdlib), designed for local open-source models...
Open-source, privacy-focused AI browser running locally on a Chromium fork with a multi-agent platform for autonomous web automation. All processing happens...
Browser-based full-process control of an AI coding agent (no client install, works on phone/iPad); 8-layer Bash security sandbox; 6-layer context compression...
Agent-first AI-native desktop shell built entirely from scratch (not a VS Code fork) on Electron + React + Monaco, where the agent is the center of gravity with...
Free, open-source AI coding agent with four-level parallelism (project, task, agent, tool), 100% local storage, offline capable, no vendor lock-in, with an...
Deep Claude integration into the Vim/Neovim workflow — chat with full visibility of open buffers, vimdiff code review, tool use (open files, run vim/shell...
First runtime self-evolving software engineering agent that treats agents as software systems which can modify their own behavior at runtime; achieves SOTA...
Local-first, evidence-first AI coding terminal with an anti-hallucination system enforcing evidence-first engineering (reading facts, verifying, distinguishing...
AI-powered coding assistant plugin for Qt Creator providing code completion, chat, quick refactoring, and project-aware tool calling for C++ and QML; dual MCP...
LLM-agnostic terminal-native AI coding agent by JetBrains; IDE/CI-CD integration; GitHub Action for auto-responding to issues/PRs/CI failures; multiple update...
AI coding agent/workspace with durable governed memory (typed, versioned documents with provenance); four-graph unified store (code + knowledge + memory +...
Java 8+ agentic SDK combining unified LLM access, Tool Calling, MCP, A2A, RAG, Agent Runtime, and a built-in Coding Agent CLI/TUI/ACP in one SDK
Formerly DeepV Code; highly customizable AI coding assistant that understands entire project context, acts as an autonomous agent with Shell/File/Web tools,...
AI platform engineering team member that works alongside you in production environments via Slack integration; built on OpenAI function-calling; covers DevOps,...
Runs the agent inside the browser itself — the live DOM replaces MCP, a WASM Debian VM replaces the shell, and agents coordinate peer-to-peer over WebRTC with...
Local-first terminal coding agent powered by Ollama; auto-selects models by measuring hardware (VRAM/RAM/cores); two-model system with escalation on failure;...
Local AI agent desktop app for macOS, Windows, and Linux that packages the agent capabilities of pi into a standalone installable application — bringing the...
Self-improving coding agent that runs an iterative loop (evaluate on benchmarks, store results, modify its own codebase, repeat), bootstrapping capabilities...
Terminal-native, pipe-composable, scrollback-safe CLI coding agent for heavy users. Inline-only (never hijacks terminal), model-agnostic, headless-first,...
AI Frontend Assistant that turns Figma designs, images, and prompts into production-ready code while maintaining design standards and coding style;...
Multi-agent code generation framework with three specialized agents (programmer, test designer, test executor); independent test case generation and iterative...
Terminal AI coding agent with context-aware code completions; summarizes full project context using Tree-sitter; maintains conversational/code context per...
Each chat thread runs its own coding agent in a Cloudflare Durable Object (not a VM) for persistent state without container overhead. Custom agent harness (not...
High-performance AI coding agent in Rust with rich terminal UI; cross-platform; extensible tool system; context export/restore persistence; OpenAI-compatible...
Multi-agent orchestration for Claude Code: Opus 4.6 decomposes tasks into a dependency graph of subtasks, Haiku worker agents execute in parallel with...
Terminal-native AI coding agent with per-request smart model routing across 6 cost tiers; real code knowledge graph (call graph, semantic search, blast-radius);...
A deliberately minimal terminal coding agent shipped as one static Go binary with a four-tool toolbox (read, write, edit, bash), no mandatory MCP, extensions in...
CLI coding agent built for any model, especially small/local models with tight context windows. Pure Python (no framework) with graduated compaction, persistent...
A complete, from-scratch agent (not a wrapper or orchestration kit) with its own compact kernel (~1,670 lines), 61 built-in tools, 77-role multi-agent fleet,...
Brings GitHub Copilot-style AI assistance natively to the Delphi IDE - a niche with few alternatives. Supports cloud APIs and fully offline AI via Ollama,...
Command-line AI coding assistant powered by DeepSeek Coder models; supports both local mode (via Ollama, free & private) and cloud mode (API key). Features code...
Educational AI coding agent CLI (~14K lines of production code) purpose-built for learning how modern coding agents work internally - a manageable scale between...
AI coding agent built specifically for game development. Speaks game-engine language (materials, blueprints, terrain, UMG, skeletal animation, packaging),...
Open-source, language-agnostic LLM-based mutation testing tool. It generates code mutants using LLMs and runs test commands to measure mutation coverage (killed...
Augment's agentic coding CLI for the terminal. Understands your codebase and helps ship faster by analyzing code, making safe edits, and automating routine...
Open-source AI agent for the terminal. Describe a task in plain English and it plans, uses tools, and delivers. Free (OpenRouter free models), local (Ollama/LM...
Stem agents that maximize cache reuse across explore/plan/execute phases. Tree-sitter virtual filesystem that minifies code for 20–50% fewer tokens....
Lightweight (<20MB), fast native code editor with AI agents, tree-sitter based code intelligence, DAP debugging, extension marketplace, ripgrep-based fast...
Open-source, local-first AI full-stack app builder with hybrid web + desktop (Electron) support. Vendor-neutral architecture with dynamic switching between 19+...
First generalist SE agent handling multiple task types (GitHub issue resolution, repo-level code generation, fault localization, program repair) across multiple...
A general-purpose Python code optimizer that uses LLMs to generate multiple optimization ideas, tests them for correctness, benchmarks for performance, and...
Open-source Claude Code alternative with hybrid architecture (Rust backend + React/Ink terminal UI), supporting both cloud and local LLMs with agentic...
Self-hosted, fully open-source autonomous SWE agent on AWS serverless infrastructure — offers a Devin/OpenAI Codex/Google Jules-like experience with no upfront...
AI-native IDE built for modernizing legacy systems, developing firmware, and migrating regulated codebases; open-source Cursor alternative with AI chat, inline...
AI-powered code generation tool for scratch development of web applications using a team collaboration of autonomous AI agents — users assemble a group of AI...
Model-agnostic, general-purpose AI coding agent that runs in containers by default; terminal-native, supports multiple isolation methods (Docker containers,...
Terminal-first AI coding agent/harness with multi-model routing; supports local (Ollama, LM Studio, MLX, llama.cpp) or cloud providers with one TUI. MCP-native;...
Minimal harness demonstrating long-running autonomous coding with the Claude Agent SDK; two-agent pattern (initializer + coding agent) with Linear as the core...
A full coding-agent harness in under 400 lines of shell and under 50KB with zero package dependencies — it runs on just sh, curl, and awk plus an API key,...
Minimalist terminal coding agent for local LLMs: three slash commands, one embedded system prompt, no router/sub-agents/skills/MCP. Designed for local LLMs...
Full-stack AI-powered SWE agent platform: Next.js dashboard, MDX docs site, terminal web client, and installable CLI coding agent ('supercode' on npm), plus a...
DeepSeek, Qwen, ChatGPT chats and coding agents in PyCharm
Minimal, transparent CLI agent framework combining Claude Code-style coding with Manus-style OS control; single-file core engine (agent.py, ~1,000 lines) with...
Multi-agent code generation framework for competitive programming using four LLM agents (retrieval, planning, coding, debugging) with an adaptive agent...
Open-source terminal AI coding agent written in Rust; 100% AI-generated codebase (human acts only as product manager). Includes MCP support, plugin marketplace,...
Open-source implementation of a Claude Code-style AI assistant built 100% with Codex (AI-assisted development traces intentionally kept in repo); implements...
Fast, self-improving terminal-native AI coding agent built with Bun and Ink. Features 40+ built-in tools, Agent Skills system with auto-generation, Code...
Restrained coding + tasks CLI with both TUI and browser Web UI sharing the same runtime; minimizes unnecessary context usage via managed compaction, lazy-loaded...
Lightweight, IDE-agnostic CLI that integrates AI code generation into any workflow by live-monitoring files in any text editor for simple //> <// prompt...
Early experimental AGI that generates a Python app from a user description. Uses LangChain and adopts the BabyAGI concept from @yoheinakajima. Very early-stage...
Open-source AI-powered terminal agent built with Go, Bubble Tea TUI, and Open Agent SDK. Features 78 built-in slash commands, custom Skills system (user-defined...
Research presenting 4 progressive levels of self-improving code agents; the Arena Loop (L4) uses adversarial co-evolution where code agents and test agents...
Agentic TUI for existing Go codebases that uses AST parsing to extract compressed, relevant context for the LLM. Two-step context approach: first extracts...
Feature-rich Claude Code clone in TypeScript (~9200 lines, 46 source files) with prompt caching (~90% input token savings), 21 built-in tools, zero external...
Open-source AI coding agent in Rust with native GUI, terminal mode, and MCP integration (both as MCP client and headless MCP server). Features transparent UI...
Next-generation AI coding assistant with three run modes (CLI terminal, Web UI, Headless JSONL for CI/sandbox). Features 20+ built-in tools, automatic...
Swift reimplementation of a Claude Code-style coding agent CLI built stage-by-stage to explore coding-agent architecture. Philosophy: coding agents benefit more...
Open-source AI app builder prototype using E2B cloud sandboxes for secure code execution; supports multiple tech stacks (Python, Next.js, Vue, Streamlit,...
Terminal coding agent running entirely on your own Cloudflare account via Workers AI; 262k context window, AI Gateway-confirmed cost tracking, local...
Semi-autonomous conversational CLI coding assistant using OpenAI function calling (no LangChain); AI proposes commands, developer approves, then auto-execution;...
Judgment-first local-first coding agent runtime where model output is treated as a proposal, never authority; durable authorization + recorded validation for...
First practical LLM-based von Neumann-architecture stored-program automatic computer; uses a self-bootstrapped prompt-program where each instruction step is...
AI coding assistant for large/complex codebases with UI context chooser for selecting files/folders and a code change planning feature.
Terminal-native AI coding assistant with hierarchical memory (user + project scopes, background memory agent), goal mode for long-running objectives, 4 built-in...
Terminal-based autonomous AI coding agent similar to Claude Code/OpenAI Codex but works with many more LLMs; uses LlamaIndex Workflow, Tree-sitter, Ripgrep, and...
Single self-contained Rust binary AI coding agent (no Node/Python runtime); 295 curated agent personas across 30 domains; built-in project wiki for...
AI agent that bumps npm dependencies and automatically fixes breaking code changes using AST analysis (ts-morph) and a plan graph DAG (based on Microsoft's...
Non-interactive design (give it a prompt, it does the work autonomously); multi-session web orchestrator (Hub) with browser UI; TUI dashboard; edit-to-fork...
Terminal-native AI runtime (not a chat interface) for controlled engineering execution. Features the IOSM methodology (Improve -> Optimize -> Shrink ->...
Browser-based AI IDE using WebContainers for real-time in-browser code execution, Monaco Editor with AI autocomplete via local Ollama LLMs, multi-stack...
Provider-neutral multi-agent control plane with a 4-step loop (Scope, Route, Verify, Replay); turns goals into bounded DAGs with owned paths; evidence-gated...
Autonomous multi-agent AI engineer in VS Code; self-healing execution loop; provider failover with cooldowns; hybrid memory/search (vector + FTS4 + MMR);...
Multi-agent LLM-based RTL (Verilog) code generator; automates register transfer level code generation using multi-agent approach; runs against...
Token-efficient context engineering that saves 70%+ tokens on long sessions by using a persistent knowledge graph (Agentic Session Memory) to recall relevant...
Python-based AI agent replicating Cursor's coding assistant capabilities; function calling with registered tools; codebase semantic search, grep, fuzzy file...
First open-source AI mobile app builder (open-source alternative to Lovable/Bolt.new/Rork); describes app in plain English and generates code in E2B cloud...
Terminal-native AI coding assistant powered by local Ollama models; 19+ specialized tools; full MCP server support (resources, tools, prompts); agent tool...
Minimalism challenge coding agent golfed down to just 493 bytes (src/smallest-agent.js); functional coding agent with unrestricted bash access; commented...
Desktop AI 'weaver' (CodeLoom) for multi-model development; orchestrates multiple open-source LLMs (Mistral, Llama, Phi) into a single fully offline desktop...
Bring-any-model via OpenRouter; six graduated approval policies with dangerous-command rejection; context pruning + history compaction for long sessions;...
Loop-based agents with success criteria (not fixed steps); forge is source of truth (no external tracker/YAML); parallel-safe git worktrees;...
It works with hundreds of tool-calling models, any model with an LLM plugin, and other Datasette plugins can register extra tools for it. With the Datasette...
4B-parameter MS-SWIFT domain expert agent distilled from Qwen3-Coder-Next; simulates a Claude Code-style agent environment; trained on 361K samples with 32K...
Single 34 MB static binary with zero runtime deps (no Node/Python/Docker/cloud), ~10 ms cold start, runs on 512 MB VPS or Raspberry Pi; verify-gate runs...
A recursive coding agent whose root never calls a tool itself: every action is a delegation to a specialist subagent, and repeated failures trigger an async...
A single Python file whose agent has exactly one tool — exec — and must grow every other capability by editing its own self/ directory of markdown and scripts...
Terminal-native and review-first: all tool calls, edits, diffs, and plans stay visible before you sign off; model freedom across many providers from a command...
DeepSeek-first coding agent with a native desktop shell (Tauri/Rust) providing a secure OS boundary for file, shell, process, Git, keychain, artifact, and MCP...
Vim-inspired philosophy-lightweight Rust build, Lua-based configuration (Neovim-style), built-in Skills system, usable as both CLI and embeddable Rust library,...
Self-hosted, on-device AI coding workspace/assistant for Android 8.0+. Runs a real tool-call loop where models can read/edit files, run shell commands (via...
AI-native architecture with 9+ agents orchestrated via YAML DAG workflows, 120 FPS rendering with Rust core, local-first privacy with edge model support,...
Open-source alternative to Vercel's v0 agent, functioning as an AI background agent for your codebase
Works with ANY LLM via Ollama local models or cloud APIs (not locked to one provider); ~95% feature parity with Claude Code for CLI use; 44 tools and 80+ slash...
Local-first (no session uploads), bring-your-own-model across 100+ providers with universal KODER_* env vars and OAuth-backed subscription logins (Google,...
C# and .NET 10-native coding agent runtime for teams building AI developer tools, agentic CLIs, and MCP-enabled workflows. Ships a terminal-first agent runtime...
Open-source AI debugging agent for VSCode inspired by SWE-agent; embeds a debugging loop directly inside VSCode for real-time developer feedback mid-run;...
Native Windows code editor built with Rust + Win32 API using Direct2D rendering and a Piece Table text buffer. Features multi-cursor editing, Tree-sitter,...
VS Code extension AI coding assistant and debugger, a mod/fork of Roo-Code (which derives from Cline). Features parallel subagents with isolated contexts, 25+...
Uncensored-no policy filter between you and your code. Does work mainstream agents refuse (pentest PoCs for patched CVEs, deobfuscation, credential rotation...
100% local-first full-stack app builder with no cloud dependency or data leakage; bring-your-own API keys (no vendor lock-in); framework-agnostic;...
VS Code extension providing a conversational AI coding assistant powered by the DeepSeek API with an agentic loop, multi-turn tool calling, and streaming output...
VS Code extension embedding an AI coding agent powered by the Pi agent runtime, with deep editor integration including inline diff previews, subagent...
Compiles the codebase into a resolved, queryable map before answering: tree-sitter grammars resolve imports, class ancestry, and every identifier to its...
Local-first AI engineering CLI implementing a 'compounding engineering' philosophy: every todo resolution automatically codifies learnings into a knowledge base...
Terminal agentic coding assistant compiled into a single 1.6MB static Rust binary (no runtime needed); drives 30+ built-in tools and supports switching between...
Composable agent runtime pairing any frontend (shell, TUI, GUI) with any agent backend over one shared extension layer. A pure kernel (typed event bus + handler...
Bito's AI Code Review Agent provides automated, context-aware code reviews in Git workflows (PR/MR) and IDEs, identifying bugs, code smells, and security...
Terminal AI developer assistant with smart project indexing (/init) that scans codebase and infers conventions; uses multiple AI models for different task tiers...
Local AI workstation combining LLM chat/roleplay, in-house Stable Diffusion image generation, voice cloning, research-grade model ELO testing, forensic...
AI agent that automates code management and feature integration in Next.js codebases (primarily the next-enterprise boilerplate) through natural language...
Agentic terminal-based coding assistant that understands your codebase and helps you code faster via natural language commands (file ops, context gathering, AI...
Desktop AI coding harness (Tauri 2) where every tool is toggleable, every description editable, every provider pluggable; agent only sees the surface area you...
Lightweight Vercel AI SDK-based coding agent library providing core primitives for building AI-powered coding assistants with tool support, session management,...
Adversarial multi-agent Code Agent System using dual-agent adversarial refinement (Test Patch Generator + Code Patch Generator) that iteratively improve;...
Dual-mode local SWE agent workbench: Assist mode (agent drafts edits, user approves each one) and Auto mode (runs bundled mini-swe-agent in isolated git...
Open-source AI agent that autonomously explores a web app URL, generates commonsense e2e test specifications, executes them, and saves passing tests as reusable...
Tiny single-binary Rust agent; persona-driven (architect, security, zen, hacker, etc.); recipe-based automation for CI/CD; multi-LLM sovereignty; safety-first...
Community fork of Google's Gemini CLI modified to work with Qwen models from Alibaba Cloud; CLI AI workflow tool with dynamic MCP server management, multi-agent...
100% Llama 2 inference with no OpenAI API keys needed; autonomously generates startup ideas, business plans, and React codebases from simple user intuition...
Open-source AI code editor (Cursor alternative) forked from Void Editor/VS Code; BYOK with any provider, local model support, subagents, plan mode, MCP...
Open-source self-hosted FOSS alternative to v0/Lovable/Bolt/Replit Agent; agent reads/writes/patches files, runs shell commands, and provisions services inside...
Multi-agent PR review + downstream fork-sync risk analysis for maintainers; combines deterministic review with multi-provider tribunal consensus, local git...
Headless coding agent built on the Embabel agent platform; aims to accelerate development with AI without using any commercial coding agents; explains code,...
AI-powered codebase navigation directly in terminal; local-first (Ollama) with privacy focus; intelligent code search via embeddings; agent mode with tool...
Evidence-gated memory system that preserves source refs across context wipe; autonomous terminal coding loop with plan/code/chat modes, custom agents, todo...
Local-first AI coding agent with a native Go core for VS Code-compatible editors and terminal; review-before-apply edits, checkpoints (workspace snapshots), and...
Local desktop-first agent workbench with a Rust runtime and Tauri/React GUI; durable sessions, approval-gated tools, subagents, goal tracking, project memory,...
Independent open-source AI coding CLI built for freedom and zero telemetry, local-first, built with Bun and rendered via a custom React Ink abstraction;...
Open-source AI agent for software testing. Uses an LLM via OpenRouter to automatically generate test plans, generate pytest test code for API endpoints, run...
Free web-based BYOK AI code editor/generator using Monaco editor with LangGraph orchestration pipeline (plan->config->generate->finalize) for AI project...
Lightweight terminal-first AI coding agent using OpenAI-compatible Chat Completions API. Built-in tools for reading files, applying precise edits, searching...
Open-source, terminal-native coding agent (IDE) keeping you in the terminal. NeoLens built-in local codebase explorer visualizes TypeScript dependency graphs,...
Small, fast, Lua-scriptable AI coding agent for the terminal (scriptable like Neovim). Custom terminal renderer (not ratatui). Built-in Vim editor (motions,...
Terminal-native AI coding agent connecting to 17+ LLM providers with beautiful terminal UI (themes, syntax highlighting, code diff previews, Mermaid diagram...
Open-source local AI-powered Flutter mobile app builder. Users build mobile apps by describing them. OSS version is a local editor for a single project declared...
Open-source debugging agent that runs locally alongside coding agents and bridges them directly to production runtime data; captures deep, unsampled full-stack...
Multi-Role AI Agent Runtime orchestrating multiple AI roles through a quality-gated pipeline (Brain -> Miranda -> Pappy -> Benson). Maestro manages a subagent...
An ultra-lightweight Codex-compatible terminal coding agent written in C++23 with no runtime dependencies beyond libcurl and OpenSSL on Linux — it authenticates...
Molecular Dynamics Agent — a fine-tuned LLM-based multi-agent system for generating, executing, and refining LAMMPS thermodynamic simulation code in materials...
AI dev overlay/proxy that lives inside your running Next.js app rather than a separate IDE or CLI; runs a proxy on :3001 in front of the dev server on :3000 and...
Fully self-hostable and air-gappable multi-tenant AI app builder with row-level security isolation, sandboxed code execution (per-project Linux UIDs, seccomp,...
Terminal AI coding agent with native real-time P2P collaboration (no cloud intermediaries), sub-500ms time-to-first-token via a custom React/Ink renderer and Go...
Self-repairing autonomous agent for digital-consciousness backup that can debug, edit, and compile its own source code under human approval; uses a unique...
Multi-agent AI code review built on Google's ADK with three specialized agents (Code Analyzer, Security Checker, Quality Reviewer) running in sequence/parallel...
Free Visual Studio extension for code refactoring, unit-test generation, documentation (Markdown/HTML), and code review using only locally-installed LLMs...
LLM plugin that adds an `llm code` command for interactive coding-agent sessions; works with any LLM-supported model (provider-agnostic) and leverages LLM's...
LLM agent that generates industrial scenes through executable C# code with precise quantitative control (measurements/positioning), structured/calculable...
Privacy-first terminal coding assistant (VibeCoder) running entirely on local Ollama models with no data sent to external services; uses DuckDuckGo for...
Unified AI vibe-coding IDE: native terminal, code editor, and coding agent in one desktop app built 100% in Rust (Tauri). Broker Agent is a one-click autonomous...
Cloud coding agent steered from the Crustocean chat platform that reads GitHub repos, writes patches, and opens PRs; stateless worker with no database,...
AI coding co-worker for APIs (code, fix, document, deploy, test); multi-agent architecture with components (coordinator, code-navigator, code-search,...
Google-free, local-friendly unofficial reimplementation of MLE-STAR. No API keys required for search (DuckDuckGo) and leverages free-tier LLMs via OpenRouter,...
Competitive multi-agent debate framework for software issue resolution. Uses competitive multi-agent debate where multiple expert agents collaborate and debate,...
Open-source BIM (Building Information Modeling) LLM Code Agent tailored for the AEC (Architecture, Engineering, Construction) industry, bridging complex BIM/IFC...
Terminal-first, local AI coding agent built in Rust (single statically-linked binary, ~20 MB idle memory, millisecond cold start). AST-powered code indexing...
100% local and private open-source terminal AI coding agent — code never leaves your machine, no cloud, no accounts, works fully offline. Free forever with no...
Local-first AI coding agent for desktop, terminal, editor (VS Code, Visual Studio, JetBrains), and CI workflows. Works inside a real local repository (not a...
Local-first autonomous coding agent (v1.0.0) that plans, edits, runs commands, debugs failures, validates results, and continues until software tasks are...
Cline/RooCode/Kilo fork for AI-assisted coding, planning, and development
Self-supervising AI code generation tool that iteratively generates, refines, compiles, and tests solutions against extracted test cases until they meet...
Fullstack Next.js app builder using Spark Engine AI's multi-agent system for interactive generation of pages, APIs, and database schemas, with automated...
Agent-native development platform spanning CLI, Web, Slack/Teams, Linear/Jira, and Mobile. Top-performing 'Droid' agent in terminal benchmarks. IDE integrations...
Local-first, plan-first, multi-agent, programmable software-engineering runtime with its own scheduler, sandbox, default-deny permission system, state machine,...
A coding agent built Android-first for Termux: PLAN mode asks requirements questions before any edits, AGENT mode runs autonomously, and a built-in live web...
Terminal-based multi-provider AI coding agent with a 'Tech-Priest of the Adeptus Mechanicus' persona and a built-in keyword-based retrieval-augmented generation...
Markets itself as the only AI agent for developing both hardware and software: it retrieves best-in-class context from the web, buys parts and subscriptions...
A native macOS coding agent written in 100% Swift with no Electron or web shell, combining project-aware chat, local tools, Git workflows, Computer Use,...
Assembles a team of AI agents that work autonomously on your GitHub repo — opening issues, debating in comments, writing code, reviewing PRs, voting on...
Agentic AI framework that autonomously modifies, analyzes, and tests codebases. Uses a multi-agent system (PlannerAgent, CoderAgent, Tester) to collaboratively...
Advanced multi-agent system built with Letta AI and DeepSeek, featuring memory optimization and Lightning AI deployment. Comprises a Research Agent, Coding...
Terminal-based AI coding agent — 14 kB package, 3-second install, smart model routing/switching, works in any terminal, supports any model. Reads, writes, edits...
Free repo-aware autonomous coding agent (DeepSeek) that edits files and runs commands
Unifies AI agent, code editor, file explorer, shell, approvals, diffs, and durable sessions in one keyboard-driven TUI workspace. Every shell command runs...
Local-first coding agent running Ollama models with tool use, diff approvals, and voice
Deliberately tiny, hackable code agent and experiment platform with a policy-blind core loop (~165 lines in src/agent.ts) where all behavior is injected via a...
AI-powered agent that autonomously deploys any GitHub repository on a user's local machine. Uses LangGraph workflow to clone, set up, and run repos with...
AI-powered coding assistant/agent for VS Code
AI coding agent for planning, building, fixing code with AI code review
AI Coding Agent for the Terminal by Fetch.ai. Binary releases repo only — no source code, just install script. Only 1 commit, minimal information available.
Open-source Claude Code–style agentic CLI built in Go with 13 built-in tools (Read, Write, Bash, Edit, Grep, Glob, AskUserQuestion, TodoWrite, WebFetch,...
Java modernization agent with guided upgrades, migration, and unit test generation
Open-source AI coding bot that automates development tasks, transforms Figma designs into deployable code, and integrates into GitHub workflows via webhooks...
A terminal coding agent built on the Pi harness that bundles goals, background jobs, and child agents in one session — plus Japanese-aware terminal layout,...
NOTE: Not a coding agent harness. Domain-specific agent that translates natural language to Infrastructure as Code (IaC) deployments with multi-LLM provider...
Provider-neutral AI coding and research agent with editable Plan Board
Autonomous AI software engineering agent ReAct loop in IntelliJ terminal
GitHub Action that uses an AI agent to write automated tests on pull requests. Built on the actions/typescript-action template. 153 commits, 1 star.
Build mobile apps by chatting with AI; one-click publish to App Store, Play Store, and web from one codebase without opening Xcode/Android Studio; built-in...
AI-powered accessibility fixing agent
AI coding agent for Apex/Salesforce development (ex Einstein for Developers)
Turn-based Salesforce agent combining Claude Agent SDK or OpenAI Agents SDK
Multi-agent AI development system for automated tool creation and testing. The GitHub repository is no longer accessible (returns 404).
JetBrains AI coding agent, chat, and code completion for VS Code
Open-source IntelliJ AI coding assistant with code generation tools
AI autofix PRs for vulnerabilities (Belgium)
Enterprise-grade multi-agent collaborative AI dev platform covering the full R&D lifecycle (requirements, task decomposition, codegen, testing, delivery) with...
AI coding sidebar for Xcode/iOS developers
Orbs — remote machines that keep working after you close your laptop; threads resumable from any device (web, terminal, phone); Puck voice control; plugin...
Auto-remediation of code vulnerabilities in CI
Google's terminal-first agentic coding CLI supporting background subagents, MCP, and multiple execution modes; part of the broader Antigravity multi-surface...
App developer agent for creating web and mobile apps
Builds mobile/web apps from a single sentence description by scaffolding features, data, and UI in minutes; allows one-click publishing to the web
Self-driving AI IDE for proactive, autonomous product development. Marketed as 'Your own coding agent, anywhere.' Website is minimal with only a hero section...
AI coding assistant offered as IDE extensions for code generation, review, and documentation. Homepage content could not be retrieved/verified during enrichment...
Private AI workstation (Electron desktop app) that plans, writes, and verifies code with every edit requiring approval; code stays on machine, works...
AI agent orchestration platform (Cosmos) running an 'always-on software factory' from ticket to code, review, verification, and deployment. Features reusable...
Agentic modernization of .NET, mainframe, and VMware workloads
AI pair programmer powered by Claude and GPT-5 (Cline fork)
Baidu's AI coding assistant (Wenxin Kuaima) with Zulu agent
Baidu's AI-powered no-code application creation and publishing platform. Hosts 189,000+ user-created apps (web apps, mini-programs, H5 games, tools). Integrates...
Full-stack AI vibe coding platform - describe apps/websites/agents in plain language. Backend, auth, payments, hosting, and SEO/GEO tools all built-in with no...
Coding agent focused on finding and fixing bugs autonomously
Datadog's AI agents for incident investigation and dev fixes
AI coding assistant with real-time code completion, documentation, and debugging suggestions
Batch autonomous engineering platform for enterprise-scale codebases
AI app and website builder by StackBlitz with Bolt Cloud for backend infrastructure (hosting, databases, auth, SEO, custom domains); auto-routes to the right...
Visual Copilot/Fusion: design-to-code and AI visual editing agents
Speed-first routing architecture that sends straightforward work to fast models and escalates to deeper reasoning only for complex tasks, finds relevant code...
Chinese agentic coding platform for Java
CatPaw is a coding agent comparable to Claude Code, distributed as a JetBrains plugin and a standalone VSCode-based IDE (per third-party references such as the...
Full multi-agent AI system that reads, writes, executes
Charlie: GitHub-native AI engineer for issues-to-PRs
AI-powered full-stack app builder ('Vibe Shipping Studio') turning natural language descriptions into production-ready deployed web applications. Full-stack...
Verifiable AI code translation for edge/defense/automotive; $125M Series B
BYOK AI coding agent delivered as IDE extensions (VS Code, JetBrains) with Agent and Planning modes plus MCP for linking external APIs/docs/databases;...
Free AI-powered code acceleration toolkit (now Windsurf Plugin) with autocomplete, chat, and search in 70+ programming languages; 3.9M+ installs; IDE-integrated...
On-device AI coding agent automating the software development flow
Agentic large-scale code migration platform (open-source core)
AI coding assistant specifically for Xcode — tab-to-autocomplete and agentic chat directly within the iOS/macOS development workflow. Supports all major LLMs...
AI-driven agentic change management with Codegraph for deterministic impact mapping, Change Stack for explaining massive diffs visually, agentic review loops...
GitLab China (JiHu) agentic coding extension
OpenAI's coding agent available in ChatGPT, as an IDE extension, and as a CLI; completes engineering tasks end-to-end (features, refactors, migrations, reviews,...
Local-first AI development OS with a multi-agent team (Prometheus, Athena, Scout, Iris, Matis, Hermes) working in parallel. Three modes: Co-Pilot (hands-on...
AI coding assistant by Sourcegraph that leverages the Sourcegraph Search API to pull deep context from local and remote codebases; available as VS Code,...
AI code assistant that uses Sourcegraph's code search to pull context from local and remote codebases; available as IDE extensions (VS Code, JetBrains, Visual...
Site shows 'Deployment Paused' — product is not currently operational. Originally described as an AI-powered IDE that teaches programming concepts in real-time...
Task-centric AI flow from GitHub issue to plan to code to PR; understands repo context and lets developers iterate on AI-proposed plans before implementation....
AI SAST with automatic vulnerability fixes
Snowflake's data-native AI coding agent CLI with 55+ built-in skills spanning SQL, data governance, dynamic tables, ML, streaming, cost intelligence, lineage,...
Text-to-app builder with 40+ integrations, code export, visual QA via computer use agent, and Max-tier parallel agents; powered by frontier AI models.
Prompt-to-web-app/landing page builder
Fully autonomous coding assistant for JetBrains
Web-based AI business-building platform with a marketplace to hire pre-assembled multi-agent AI teams (Engineering Team, Marketing Agency, etc.) with distinct...
AI-powered coding agent and security scanner
Highly non-developer-friendly AI app builder - users describe their app in natural language and collaborate with an AI agent to iterate. 1-click deploy to...
AI engineer in dbt for models, tests, docs
Terminal coding agent for DeepSeek V4 models. Streams reasoning blocks, edits local workspaces with approval gates, reads/edits files, runs shell commands,...
AI-powered website builder on Hugging Face Spaces that uses open-source models (DeepSeek, MiniMax, Kimi). Combines AI code generation with free hosting, instant...
AI agent that generates unit tests for complete code coverage across a project, requiring only your repository as input
Dev Agent is an AI-powered intelligent development assistant plugin for IntelliJ Platform
AI-powered programming assistant built into Huawei's DevEco Studio for HarmonyOS development. Understands project context for ArkTS code generation/completion,...
Autonomous AI software engineer by Cognition with a workspace (shell, IDE, browser); accessible via web, CLI, Slack, Microsoft Teams, and API; features MCP and...
GitHub-native AI developer agent for issues and PRs
Open-source AI coding assistant supporting local and cloud LLMs
Built for fast notebook generation; upload data files (.xlsx, .xls, .csv, .json - up to 5 files) to create an analysis and generate a Python notebook
Mistral Vibe CLI is a native terminal-based coding agent with MCP server support (HTTP, streamable-HTTP, stdio), a Skills plugin system with custom slash...
Was described as an AI-powered web development agent for prompting, running, editing, and deploying full-stack applications. GitHub repo (RaheesAhmed/devstream)...
Transforms natural language prompts into fully functional, production-ready web and mobile apps without requiring coding experience or drag-and-drop interfaces....
Marketplace to 'hire' pre-configured, multi-agent AI teams (e.g., full-stack Engineering Team with Technical Lead, Backend Engineer, Frontend Engineer, QA...
Erdos was a Jupyter-first data science IDE from Lotas with an AI assistant that read and wrote notebooks, Python, R, and Quarto directly — grounded by a...
Agent that adapts itself in real-time based on the task at hand, using pre-defined agent personas tailored to specific domains. (Domain evo.ninja is dead and...
IDE-native AI agent using JetBrains IDE APIs instead of terminal commands
Exponent ran as both an interactive pair-programming session and a fully autonomous background worker triggered from Slack or GitHub, blurring the line between...
Enterprise-grade 'autonomy stack' for software development with a customizable harness supporting MCP, hooks, and subagents. Model independence as a platform...
Prompt-to-app vibe coding platform
AI-powered prompt-to-code prototyping within Figma. Uses design context (Make kits, npm packages, Figma frames, PDFs, attachments). Code-backed and visually...
Cloud-based agentic IDE in the browser for building full-stack AI apps (APIs, backends, frontends, mobile), powered by Gemini; built on Code OSS with Open VSX...
Coding agent for Android Studio with emulator and LSP access
AI-powered coding assistant plugin by Fitten Tech (Beijing) for VS Code and JetBrains IDEs, focused on code completion and chat. Could not verify details from...
Y Combinator-backed platform that turns Claude or ChatGPT into a full app builder with backend, database, hosting, logins, and one-click publishing; available...
1M token context window, enterprise security and IP indemnification, code customization using private codebases, deep integration with Google Cloud services...
Cosine ships the 'cos' CLI powered by its own Lumen models (Scout post-trained from Devstral 123B, Outpost from Kimi K2.6, Sovereign coming) with multi-agent...
Genspark is an all-in-one AI workspace with related products including GenTerminal, Figma to Code, AI App Builder, and AI Website Builder. No specific 'Genspark...
Code-change automation agents for migrations, cleanup, and PRs
Natively built into GitHub; world's most widely adopted AI developer tool; integrates across editor, CLI, GitHub.com, and mobile; supports multiple leading LLMs...
GitLab Orbit lifecycle context graph connecting code, merge requests, pipelines, and deployments for agent grounding; AI Catalog for agent/flow management;...
AI coding agent for VS Code with BUILD (codegen, project setup, deployment), ASK (chat, code explanation, web search), and TEST (automated test...
Full-stack vibe coding with Antigravity agent + built-in Firebase (AI Studio 2.0, 2026)
Graph index of entire codebase; swarm of parallel agents reviewing beyond the diff; learns team coding standards from PR comments over time; TREX autonomously...
Enterprise agentic IDE plugin for Guidewire development
Hosted no-code AI web app builder
Huawei Cloud's AI-native intelligent coding product with code generation, R&D knowledge Q&A, and unit test generation. Supports HarmonyOS development. Product...
JetBrains edition of the HuiCai AI assistant with agent core
Chinese AI programming assistant with completion, debugging, project creation
Free Chinese terminal coding agent
AI coding assistant running parallel multi-agent workflows
AI-powered workspace by Augment Code where human teams and AI agents collaborate to plan and build software. Unifies code (repos, files, branches, diffs),...
AI agent that generates, runs, and fixes Java unit tests
Deep integration of AI features with code understanding and core IDE workflows. Ecosystem includes AI Assistant (assistive), Junie (autonomous coding agent),...
JD.com's commercial coding agent (joycode.jd.com) — the externalized product of JD's internal JoyCoder assistant, reportedly invoked over a million times by JD...
The name JD.com used for its in-house LLM coding assistant — IDE plugin across JetBrains and VS Code with code prediction, comment-to-code, unit test generation...
Async coding agent by Google that clones your repo to a Cloud VM, builds a plan using Gemini models, makes code edits, and opens pull requests for approval....
LLM-agnostic coding agent powered by IntelliJ IDEA Engine (top performer on SWE-Rebench); model-agnostic BYOK with cost-efficient 'plan on Opus, implement on...
AI-powered coding assistant built into JetBrains IDEs
GenAI-native end-to-end software testing agent that plans, authors, runs, and self-heals tests using natural language across web, mobile, UI, API, database,...
Terminal-based AI coding agent and shell by Moonshot AI that can read/edit code, execute shell commands, search/fetch web pages, and autonomously plan actions....
Russian AI service for developers (AI coding assistant)
Frontend-specialized coding agent for complex UI codebases
AI bug-fixing agent from runtime error snapshots (Czechia)
Agentic coding assistant with local LLM via Ollama
AI developer agent that builds production-grade apps and deploys them to your own cloud. (Website is JS-rendered; could not verify detailed fields.)
The only AI pair programmer scoped exclusively to backend API development: feature-based chats per endpoint, tab-to-prompt generation, two-way GitHub sync on a...
Language-Integrated Smart Agent AI companion
Chat agent that can read, write, and edit project files
Local LLM-assisted code completion using llama.cpp
Local-first AI coding assistant
Integrates the Llmrix multi-agent AI platform into JetBrains
Open-source infrastructure by Rubric Labs for running natural language workflows on your codebase; auto-labels, assigns, comments on, and reviews issues/PRs via...
General-purpose AI agent platform that can build websites, create slides, design graphics, create games, generate AI images and music, conduct research, and...
AI-driven mainframe rewrite platform (Imogen); Gartner Cool Vendor
No-code/low-code platform (rebranded to Workshop AI by Atlas Futures, Inc.) for building websites, AI-powered apps, dashboards, and internal tools by chatting...
Cloud-native coding agent CLI for managing remote Mentat agents from your terminal; auto-detects repo/branch context. Could not verify detailed fields —...
Lightweight AI coding agent for terminal
Auto-fixes SAST findings across scanners (Israel)
AI legacy code modernization at scale
Mass-scale automated refactoring (OpenRewrite core) + Moddy agent
Autonomous DevOps agent working inside the IDE
A coding agent purpose-built for Moore Threads' MUSA GPU ecosystem: TUI and desktop form factors, natural-language MUSA kernel and TileLang operator generation,...
AI code assistant optimized for Microchip embedded products
Proactive AI agent for writing, coding, research, and data analysis
Kuaishou intelligent coding tool with agent mode, code continuation, edit prediction
Coding agent with configurable LLM endpoint
Agent-native platform ('Agent Runners') to build/deploy web apps from prompts
AI AppSec engineer for finding/fixing vulns
Fully autonomous BYO-model AI coding assistant
Gitpod rebranded 2025 into parallel cloud software engineering agents
AI code assistant across all JetBrains IDEs
The referenced GitHub repo (github.com/otto-ai/otto) returns 404 — the repository does not exist or has been deleted. There are unrelated community repos named...
AI engineering department with autonomous Frontend/Backend (soon QA) engineers that read your codebase, auto-pull scoped tasks from your backlog, run in...
A trainable agentic framework with 'in-the-flow' optimization using Flow-GRPO RL algorithm that optimizes the planner agent live inside the multi-turn...
Mimics Copilot Chat agent functionality using Perplexity Pro as backend
Web-based AI trading toolkit including an AI coding agent that generates trading code (Pine Script, MQL5, cTrader C#, NinjaScript, thinkScript) plus a finance...
Automated code-hardening/remediation agent (pixeebot)
AI agent for FiveM/GTA5 development writing production Lua and auto-fixing errors
Combines AI builder and cloud runtime in a single product; unique 'rewind' feature rolls back entire app (including database and files) to a saved snapshot in...
AI that connects product defects to code changes and fixes
A proprietary local-first coding agent that runs as a background daemon, so the terminal UI and any protocol client are only front ends onto the same session....
Builds open-weight agentic coding foundation models (Laguna) and ships a desktop app, CLI, and agent runtime (research preview); models available via OpenRouter...
Turns codebase and SDLC into a living context graph for AI agents - indexes code, structure, decisions, source history, team knowledge, and engineering...
Was described as an advanced multi-agent AI template combining Groq's speed with LangChain's flexibility. GitHub repo (Hams-Ollo/Project-S.O.C.R.A.T.E.S.) is...
AI agent that writes/updates infrastructure-as-code
All-in-one AI dev platform (GPT Pilot core) from prompt to deployed app
Open-source AI coding agent (Chinese security vendor)
Next-generation CLI AI coding agent based on Qwen programming large model
Agentic coding platform with multi-agent expert collaboration, persistent memory & rules, skills/plugins, codebase 'wikilization' (architecture discovery), up...
Deterministic AI coding agent in secure sandboxes
Agentic coding assistant: plan, implement, review, refactor
Open-source AI agent that autonomously handles engineering tasks end-to-end
Lightweight AI coding assistant connecting directly to LLM APIs
AI app builder for creating web applications from natural language
AI mobile app builder: create native iOS apps by chatting with AI and publish directly to the App Store. Rork Max App lets users build and ship iOS apps from a...
Atlassian coding agent (CLI + Jira/Bitbucket integration) built on Teamwork Graph
Agent that clones and rebuilds websites from a URL/prompt
AI web app builder similar to v0 and Bolt — generates web applications from natural language descriptions. Website returned 429/404 during verification, so...
Joule developer agents for ABAP/SAP Build Code
Persistent Systems generative AI software engineering platform
AI coding assistant / code copilot for JetBrains
AI debugging agent: RCA from telemetry + fix PRs + PR review
Lightweight AI coding platform with agent, copilot, debugger, code generation across many models
Users can choose from 12+ AI models, get managed Supabase included, retain full GitHub export and commercial rights to their code/database/hosting, and deploy...
Repo-native autonomous software engineering platform
Autonomous monitoring platform that triages production alerts, removes duplicates, root-causes bugs with 84% accuracy using a codebase graph, and posts analysis...
iFlytek's design-to-code agent for converting designs to code. Homepage (iflytek.com) did not respond to fetches; could not verify further details.
China Telecom AI programming assistant
AI DevOps agent for infrastructure code
AI-driven software development framework heavily influenced by GPT Pilot and Microsoft's AutoDev paper. The GitHub repository is no longer accessible (returns...
Cursor-for-VSCode plus Cline/Roo features combined: AI peer programmer
Enterprise Context Engine that maps organizational dependencies and architecture; strong focus on security, privacy (zero data retention), and IP protection...
Intelligent coding assistant by Alibaba Cloud with an autonomous Programming Agent mode (self-planning, task decomposition, tool usage, terminal command...
AI assistant that writes and deploys server vals/apps
AI website/app builder with agent workflow
Project manager delegates coding across role agents on different models with receipts
Agentic-by-default AI app builder — plans, creates tasks, and connects to databases automatically; one-click deploy to Vercel; GitHub repo sync; design mode for...
All-in-one coding agent orchestrating top-tier AI models; high SWE-bench Verified score
Figma's Visual Copilot (Figma-to-Code plugin) converts Figma designs into code. Figma page returned HTTP 403; details could not be directly verified from the...
Built-in agent-first experience in VS Code with dedicated Agents window, multiple chat surfaces (Agents window, Chat view, inline chat, Quick Chat). Full MCP...
The Warp Agent shipped as a standalone CLI that is also a native terminal multiplexer: agent sessions run atop managed pty connections, so you can switch...
Enterprise Java modernization with generative AI
Imaginary-web app generator turned consumer vibe-coding site
AI-assisted development with specialized AI agents
AI-first IDE (now rebranded as Devin Desktop by Cognition AI) with built-in agent harness (Devin Local/Cascade) that can chat, write code, run code, generate...
AI IDE (VS Code fork, now branded Devin Desktop by Cognition) with the Cascade agent panel, Devin Local Agent, context awareness, app deploys, workflows, and...
AI agentic coding assistant working with Anthropic, OpenAI, Azure and Ollama
Analyzes your codebase to produce personalized production-ready code
Tencent's AI-native desktop agent workbench for office automation, code generation, and logic optimization. The product page returned minimal content during...
Xcode 26 built-in coding agent with ChatGPT/Claude backends and Swift Assist
BYO-model AI coding assistant with agentic multi-file coder
Vibe-coding web app/site builder from Chinese team
A self-evolving coding agent CLI that reads its own source code, assesses itself, makes improvements, and commits them if tests pass. Runs this cycle...
AI coding orchestration platform with multi-model orchestration (Claude Opus for planning, Gemini for building, OpenAI Codex for reviewing), spec-before-build...
AI-native scanning + verified security patches; RSAC 2026 Innovation Sandbox
Chinese AI coding partner with automatic code generation
AI coding agent API