Deterministic prose checks for clearer coding-agent responses
Open-source, local-first AI agent for coding and real work. BYOK models, MCP, skills, plugins, workflows, and private knowledge bases.
Swift Pieces provides a library of designed SwiftUI components featuring integrated motion, haptics, and states. Developers can easily integrate these elements via a single Swift file, CLI tool, or AI-assisted installation.
Every coding agent in one desktop workspace. Run Claude Code, OpenAI Codex, Cursor Agent and Grok side by side. Bring your own CLI. Free during beta. Open source as of 21 September 2026 (MIT).
Configure and run existing coding-agent CLIs from one desktop app.
Portable, verifiable work state for coding agents
Local-first AI agent trace workbench. Inspect Claude Code, Codex and OTLP logs, compare runs, and export privacy-conscious reports. No backend or API keys.
Local-first MCP plugin for continuous software-quality review by AI coding agents, powered by Jev.
Second-model AI auto-review for DeepSeek Harness approval requests: a read-only reviewer subagent returns structured allow/deny verdicts with reasons, fail-closed by default, fully auditable from the session log (approval/asked -> autoReview/verdict -> approval/decided).
Ten Engines. One local-first Harness OS.
Verifiable research-report engine for DeepSeek Harness: content-addressed evidence ledger (claim-snapshot binding, tamper-evident) plus versioned sealed reports with per-claim verification verdicts and a manifest-sealed directory.
A local-first, Pi-powered AI agent workspace for Desktop, WebUI, and CLI
lnwjud — local AI-agent runtime & MCP gateway
Multi-agent task orchestration for coding agents. Turn one goal into an ordered plan of tasks (each with its own runner, model and mode), then execute and verify the results.
A local desktop control center for organizing Codex projects, engineering tasks, sessions, Git changes, Skills, and MCP configuration.
You write the specs. It clears the board: coding agent chains plan, review, implement, verify and merge unattended on your own machine, on the GPT and Claude subscriptions you already have.
A persistent incremental diff review loop for pi
Where you direct your coding agents: start a job with a chat, watch them work in sandboxes, and approve what matters with the evidence.
手机平板远程 DeepSeek Harness:扫码即可继续使用电脑上的 DSH,无需远程桌面 / SSH / 公网 IP,支持一次性配对、Github授权加密登录,独立设备授权与随时撤销,实现远程连接很简单,但安全才是我们所想要的。
Everyday Git, made delightful. A modern, fast, beautiful open-source Git client for macOS, Windows, and Linux.
Local-first observability CLI for Codex and Claude Code
Open-source Documentation for Developers
Local-first live command center for Grok Build — runtime, ACP control, session history, and Git changes.
Click any React UI element, jump to its exact JSX/TSX source, inspect proven API data flow, and prepare reviewable AI patches. Local-first and development-only.
jevmem saves the decisions, rules and failed approaches from your Claude Code chats to JEVMEM.md in your repo, and brings the relevant ones back next session.
The spreadsheet framework built for agents — write the model as React, export a live .xlsx. The third medium, after open-slide and open-doc.
One Developer. An Entire Fleet of Autonomous Coding Agents. Local-first desktop agent workbench built on Electron + pi.
Claude Code-style declarative permission rules for DeepSeek Harness: ordered allow/deny/ask rules with tool-name, argument (glob/regex), and workspace-path matching on the tools/pre-execute waterfall, session-log audit, and HMR reload.
Cheaper and better Greptile alternative runs on your own github actions.
One control plane for parallel React Native work on macOS: slimmed iOS simulators & Android emulators, live dashboard, leased Metro ports, shared native build cache, and Claude Code / Codex device attribution.
Normify · DSH 插件:把项目架构写成归一化的分形模块树,三层校验(写时 / 校验 / 冻结回执)、30 个 normify_* 工具 + normify-gen 技能,一键渲染单文件交互式架构图;支持伴随式开发(change_open → brief → check → 实施 → refresh → change_close)。
CLI that picks Cursor, Claude Code, Codex, or OpenCode + model/effort for a task, then launches it. Powered by Jev and Herdr
AI-powered CLI for reviewing GitHub pull requests, detecting potential bugs, security risks, regressions, and missing tests, with structured Markdown reports for open-source maintainers.
Remote notifications for the Pi coding agent — Pushover, ntfy, and webhooks when tasks finish or need your input.
Jev-powered semantic code search for coding agents — find behavior across repositories via CLI or MCP, with exact source excerpts and line numbers.
Open-source starter stack for Agent product developers: desktop client, control plane, workers, multi-agent workflows, admin, mobile, and operations.
Codex-verified, Claude-compatible durable local runtime for Claude Code Workflow JS, with resume, checkpoints, budgets, and Git worktrees
Unofficial Factory Droid integration for VS Code and Cursor, with coding chat, code review, BYOK models, and independent code completion.
Autonomous backend platform. Your coding agent builds it over MCP, Backenly keeps it running: Postgres, REST APIs, auth, storage, realtime, and functions, with every change governed, verified, and reversible.
One shared project for your team's Claude Code, Codex, Cursor and OpenCode agents: a task board, decisions and handoffs. Everyone keeps their own plan.
A modern, full-stack social networking platform designed specifically for software developers, creators, and engineers to share projects, stories, and ideas.
The review companion that knows what you look for: ranks the PRs waiting on you, shows what the diff doesn't, and learns how you review. Runs on your machine.
Codex-native engineering workflows derived from pstack, with 45 explicit skills and 23 Poteto Mode playbooks.
Open-source desktop app and VS Code extension for Meta's Muse Code CLI, with threads, inline diffs and approvals on Windows, macOS and Linux. Unofficial, not affiliated with Meta.
🌊 32 free curated AI models (up to 1M context) for Oh My Pi & Pi. Thin by design — model list, dumb relay pool, log — add your own relays to spread requests across more IPs.
A calm agent multiplexer with familiar IDE features, for macOS, with an iPhone app.
An open-source workspace for coding agents, running on your own machine. More work. Less AI waste. Same subscriptions.
Skilldex is a local-first desktop dashboard for developers to discover, organize, and favourite their agent skills across global, project, and repo sources
Record your screen, narrate the bug, and hand your AI coding agent structured Markdown - one annotated screenshot per mark. Desktop app, CLI, and MCP server for Claude Code, Cursor, Codex, and Windsurf. On-device Whisper transcription, local-first, MIT.
Projects generated with GPT-6 Astra: web apps, games, creative tools, and developer utilities.
Remote-first tmux IDE for coding agents. Agent notifications on desk and phone, voice replies, screenshot paste over SSH, multi-host, ports, files, and Git.
A coding agent harness built specifically for Motif-3
Local-only web GUI for inspecting agent skills (SKILL.md) across user, project, plugin, cache, and marketplace sources
expose localhost to the internet instantly. zero-dependency tunneling engine with a custom 9-byte binary multiplexer, interactive bubble tea tui, and real-time request replay.
Angular DevTools : inspect components, signals, DI, and routes. Runs as a Chrome extension, embedded panel, standalone CLI, static report, or MCP server for coding agents. Built with Devframe.
A curated directory of open-source projects built by Algerian developers. Discover, star, and contribute. 🇩🇿
A free, open-source UI library for web designers and developers. Copy, paste, and customize beautifully animated components for your next project.
Natural language Jev-powered AppStore indexing tool: describe an ambiguous feature of an app, get and research matching apps from the US top-charts. Ideal for solo developers doing market research.
Finds the claims in your CLAUDE.md / AGENTS.md / skills that your code no longer supports. Zero-config, no API key.
One window for a whole swarm of AI coding agents — split terminals, per-agent git worktrees, live web previews, and a War Room where agents collaborate