Havuz

Antidote for fluffy specs, a toolkit for fact-driven development with AI agents

Compartment is a self-hosted deployment system for small software on infrastructure your team controls.

Carbon Code 是一款类似 Claude Code 的代码开发工具，也是中国第一个基于 DeepSeek 的代码开发工具，Token 成本节省 90% 以上。它可以实现自动任务拆分、自动开发、MCP 测试，以及与 Claude、Codex 等多 Agent 协作，能力接近

Local-first MCP tools and dashboard for investigating Codex token usage, credits, costs, caching, and thread patterns.

You write the specs. It clears the board: coding agent chains plan, review, implement, verify and merge unattended on your own machine, on t

Ten Engines. One local-first Harness OS.

个人 Pi 编程智能体配置：17 插件 + 18 全局技能 + 2 MCP 与一键安装脚本 / Pi coding agent full configuration

A local desktop control center for organizing Codex projects, engineering tasks, sessions, Git changes, Skills, and MCP configuration.

A Mac vault that runs authenticated actions for AI agents over MCP. The agent gets the operation, never the key: no command reveals a stored

OKF (Open Knowledge Format): Durable, structured memory for AI agents. Author, validate, consume, and maintain portable knowledge bundles th

One file. Under 200 lines. Zero dependencies. It's a coding agent.

Zero-integration Chrome extension that maps React DOM elements to JSX/TSX source locations.

Native, local-first knowledge graph engine for code and project artifacts—inspired by Graphify, built in Rust, and evolving beyond it.

Full-lifecycle wiring between Claude Code and the Clawd on Desk pet — all 15 hook events, with permission prompts answered straight from the

A persistent incremental diff review loop for pi

Boring Is All You Need: a Codex Desktop workflow replacement with transactional migration, audit-only managed hooks, risk-based review, scop

GitHub contribution automation and scheduling tool for developers. Build, manage, and schedule your GitHub activity with a simple, intuitive

Model-neutral AI workspace for macOS with multi-provider models, extensible tools, and inspectable deliverables.

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。

macOS-first local MCP and Actions runtime that gives ChatGPT real coding hands over trusted projects

Find the plaintext secrets on your Mac and move them behind Touch ID, injected just in time without breaking the tools that read them. Free

Your pocket agent. Local-first AI agents on iOS and Android — real workspaces, tool execution with approvals, and your choice of model (DSH

Route complex Codex tasks to model-specific background workers with bounded concurrency and lead-agent verification.

CodeWiki is a knowledge platform that analyzes repositories into AST graphs, builds GraphRAG indexes, and generates source-grounded develope

Persistent memory plugin for OpenCode. Obsidian-style knowledge base that survives across sessions

A Claude skill that keeps your AI coding tools from contradicting each other. Generates requirements.md, design.md, and tasks.md before an

A practical Codex skill for evidence-backed repository health audits

Project memory for long-running AI work across agents, models, and sessions. Keep context, decisions, progress, and next steps in local proj

Uncensored AI coding agent in your terminal. Does the work, skips the sermon. 256K context, crypto-only billing, no card on file.

Local companion launcher and maintenance toolkit for the Codex desktop app.

Open-source AI coding agent for your terminal — one Rust binary, 15+ LLM providers, 1,000+ models. A self-hostable Claude Code / Codex CLI a

Terminal-first, knowledge-grounded multi-agent software delivery pipeline: scope requirements, implement changes, run tests, and gate pull r

Trusted localhost HTTPS — local CA, /etc/hosts, mDNS LAN sharing, reverse proxy. Maps https://name.local → localhost:port

The context spine that 10x's every AI coding session. Live in 8 IDEs (Claude Code, Cursor, Cline, Continue, Aider, Codex, Windsurf, Zed) via

Claude Code skill for automating iOS App Store submission.

Audits AI-generated frontend and rewrites it so it stops looking AI-made: purple-gradient slop and the "tasteful" defaults (cream + terracot

Deletion-first Agent Skill for removing test bloat, verification theater, and speculative fallbacks while preserving behavior.

GAN-style self-improvement loop for any text artifact: mutate, grade with a SEPARATE model, keep only verified wins (pairwise-judged), rever

Open, MIT context format for AI coding agents: centralized .aigx/ rules + a per-file boundary index that targets the right rules, forbidden

A scrolling terminal multiplexer that tracks your AI coding agents.

Local-first observability CLI for Codex and Claude Code

Claude Code usage governor: compact professional output, context slimming, tool-output filtering, telemetry, and drift guardrails.

Measure what your brain is losing while AI does your work - unaided-skill baselines, decay curves, reliance drills

A CLAUDE.md built on Karpathy's principles and gstack, adding a verification harness and autonomous-loop rules for Claude Code.

Goal-based planning and proof gate for AI coding agents

Persistent memory plugin for Codex. Obsidian-style knowledge base that survives across sessions

Automatically completes the full workflow from requirement research → research review → planning → plan review → development → development

your repository as generative art — npx codebase-posters

Claude Code, Codex and Pi write down everything you do. bough draws it and prices it: a day, the work inside it, every prompt, and what each

A drop-in proxy that compresses bloated code context in real-time, cutting LLM API costs by 50–80% without losing what the model actually ne