Havuz

DeepSeek ve uyumlu modellerle terminal, desktop, browser veya ACP destekli editör üzerinden kullanılabilen, uzun süre çalışan bir coding agent sunuyor.

Neden yükseliyor · 159 günde 35.720 yıldız ve son 28 günde 932 commit dikkat çekici ilgi ve geliştirme temposuna işaret ediyor; ancak aynı dönemde yalnızca 2 farklı yazar görünmesi, aktivitenin geniş bir katkıcı tabanına yayıldığını göstermiyor.

Mods for Claude Code. effortless picks the reasoning effort for every prompt and shows the cache countdown. textless (desktop app) shows art

Claude Code mod: hands off to a fresh conversation when context fills up, keeps the cache warm while you're away, and turns what you teach C

Context compression for AI coding agents: compresses tool output before it enters the model, dedups repeats to 13-token refs. Claude Code, C

Hermes Agent context engine: the main model writes the compaction handoff on its cached prompt prefix

Fast, tiny AI coding agent for the terminal, written in Rust. One small native binary for Apple Silicon Macs. Works with OpenAI, Anthropic a

A tiny Claude Code skill that keeps your prompt cache warm during idle sessions, so your next message reads from cache instead of paying ful

The one place for the skills, keys and memory of every agent you use. Formerly OMNI.

An experiment-driven LLM engineering lab with 9 reproducible Python experiments exploring token economics, API costs, caching, context growt

Build smarter. Work superior. An opinionated rule set for AI coding agents to create original interfaces, compelling content, and production

⚡ Terminal HUD, hard budget breakers & prompt-cache leak alerts for AI coding agents (Claude Code, Cursor, Codex)

One Go binary that lets a client talk to an AI contact the way it would talk to a person: JIT threads, JIT prompt cache, JIT REPL environmen

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

Claude Mods and the tools to build them: a builder skill, then mods

Multi-Agent Code Review Cost Benchmark 2026: Librarian vs Prompt Cache

Independent research on Claude Code internals, Claude Agent SDK, and related tooling.

Claude Code mod: context in tokens, 5-hour and weekly limits vs. the clock, prompt cache countdown, session cost and running agents, in one

Drop-in prompt-caching fixes for the LLM agent harness you use. Point your AI coding agent at this repo and it ships the patches.

Claude Code plugin: automatic Claude model routing. Picks Haiku, Sonnet or Opus and the reasoning effort per prompt, step and subagent; effo

Local-first AI companion daemon — an agent harness with its own long-term memory. Remembers you, grows with you, extends and repairs itself.

A cache-aware attention and forgetting layer for high-frequency everyday AI chat.

Context optimization for AI coding agents — lower token cost and latency without sacrificing accuracy.

DeepSeek Harness plugin: DietrichGebert/ponytail lazy senior mode & 7-rung ladder port, 6 skills with bilingual descriptions & per-skill tog

Claude Opus 5.5 with the effort level re-decided every step by the TypeSafe Jev reflex — without breaking the prompt cache. CLI + Claude Cod

🧠 Native macOS menu-bar dashboard for AI coding: Claude / Codex / Gemini / Grok / Kimi quotas and official Coding Plan usage (Volcengine, Al

A minimal coding-agent harness built to measure what context management actually costs: constraint survival after compaction (8%→83%), def

A Complete faster runtime for coding agents. Optimize the journey, not the hop. Make coding agents 25% faster and 30% cheaper on average whi

开源桌面 AI Agent 与 CLI：长期记忆、多 Agent、实时语音；Qwen3.8-27B Coder390 EfficientThink 本地 GGUF（llama.cpp Q2/Q3，内置 Q4 MTP）· Open-source desktop AI agent &

TypeSafe (Jev) skill routing for Hermes Agent: names the one skill worth loading, before the model call. Opt-in, stdlib only, ~$0.001 per ro

Per-conversation x-opencode-session injection for OpenCode Go in DeepSeek Harness - fixes 400 MissingSessionID without breaking other provid

Anthropic API reverse proxy with prompt-cache injection and request body transforms

Track what every AI coding agent costs and see the limit coming. Per-session, per-model cost for Claude Code, Codex, Cursor and 37 more tool

The open engine behind Aedelgard: a persistent, self-hosted AI agent with a verbatim memory palace and a model-agnostic brain. Swap Claude,

Run Claude Code and Codex across multiple subscription accounts. A local Rust proxy that sends each request to the account with quota left,

Lossless context folding for the Pi coding agent: only the oldest context changes; the fresh window stays untouched.

Compression with a quality contract — cache-aware, causally-pruned LLM context compression for agentic runtimes, certified non-inferior acro

Run Claude Code with OpenRouter models while keeping tools, file editing, bash, MCP, VS Code support, smart caching, model switching, and sa