Havuz

Cut context bloat in your AI-agent stack: find and safely prune unused skills, MCP servers and subagents from real transcript evidence

Carves your agent's tool outputs down to what matters. Never cuts what doesn't come back.

CLI output compressor for Claude Code. Reduces token consumption by 50–90% by compressing verbose command output before it enters the con

MCP-native code retrieval for AI agents — 84-88% fewer read tokens, BM25F + semantic search, AST chunks, session dedup

Cut Claude Code costs with sourced levers -- prompt caching, model routing, Batch -- plus an installable skill, CLI tools, a cost calculator

Eco mode for Claude Code. /eco: -31% to -73% output tokens with critical findings intact; /eco-max: up to -75% with lowered effort. Measured

Model-agnostic output control for LLM agents: concise, action-first answers with proof and risk preserved. Native install paths for Claude C

A lossless token-optimization codec for LLM agents — cut input tokens up to 82%, nothing dropped, every fact kept. Drop-in proxy + one-line

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

LLM-ready reasoning surface over logs

TOMC Memory: CPU context compression and task memory for Codex, Claude Code, Cursor, TRAE, WorkBuddy and other MCP clients.

CDAF (Cached Descriptive Asset Files) - open sidecar format for video so AI agents stop re-analyzing the same footage. Spec, CLI, agent skil

Fast, AI-agent-native code search in Rust — hybrid BM25 + semantic, Tree-sitter AST chunking, dependency & impact analysis. Drop-in replacem

面向 AI Agent 的高并发、低延迟上下文治理与 Prompt Cache 守护网关 (Tree-sitter AST / Tool Delta / 50%~80% Token 削减)

Context compression for MCP. Same upstream call, exact results recoverable.

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

Claude Code skill — delegate coding tasks to Mistral Vibe, supervise via git diff & save 50-90% Claude token usage

Private, provider-neutral context compression for LLMs. Reversibly compress schemas and logs inside your perimeter with zero third-party cal

Pith is the hook that makes Claude Code sessions last 3x longer.

Turn Google Drive PDFs into Obsidian wiki notes via NotebookLM MCP without loading full PDFs into Claude context

Delta MCP is a free app that sits between your AI apps and their MCP servers: any server, every tool. It checks each task before it runs, ap

Zed extension for Headroom — context compression for AI agents

EGC gives every AI coding agent the same brain. Shared memory, skills, and live context across Cursor, Claude Code, Copilot, Aider, and 29 A

State aware knowledge compression, ingestion, and hybrid retrieval engine. Zero dependencies. Sub-100ms queries.

The AI-native wire format for structured data. 100% comprehension on every frontier model. 50-92% fewer tokens than JSON. 43B+ lossless roun

Measured, monotonic, lossless input savings for Codex.

Quieter sessions for Claude Code: less narration, shorter tool output and concise answers focused on the result.

Local Claude Code and Codex plugin that bounds oversized and repeated tool output, preserving complete redacted artifacts. Runs locally, mak

Token-budget-aware graph navigation for AI coding agents. Serve exactly the noodles your LLM needs. 🍜

Keep Claude Code context clean. Open-source toolkit: drift detection, re-read dedup, integrity scoring, AST-aware reads, 15 MCP tools. 62.6%

Audit and shrink your Claude Code startup context. Measures what every skill, plugin, agent, and memory file costs in the system prompt, the

Local CLI and MCP server for compacting test and build output, with retrieval from a redacted local cache.

token-ninja routes deterministic shell commands locally — zero LLM calls, ~19µs latency. Works silently inside AI tools via MCP.

TERSE - a hierarchical state language for humans and AI agents.

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

MCP server for Unity Editor — 160 tools for scene, assets, animation, VFX, playtest & more

A compiler for AI coding tasks: pre-resolves repo facts (callers, git history, blast radius, bug-site source) before your coding agent's fir

A tool that replaces everything that can be deterministic in repository work: search, symbols, callers, task scope, history, review and Git

Codex Cost Router Skills - preflight task routing skills for Codex to reduce token waste, choose model tiers, reasoning effort, permissions,

MCP server that shrinks logs, JSON, docs and web pages before your AI reads them. Offline, deterministic, ~3ms, no API key. 66% fewer tokens

Chrome Extension that lets you continue any AI conversation anywhere—without losing context. No more copy-paste—move full context across AI

Token-compression skill. An adaptation of caveman — short common words, trust context, say just enough, be laconic.

Verdict-first output for AI coding agents. Tiny prompt + installer for Claude Code, Codex, Gemini, Cursor, opencode, and 30+ agents.

Agentic coding framework and meta-harness for planning, building, reviewing, and tuning software across coding agents. Portable agent squads

Guardian Runtime sits securely on your machine, acting as a bodyguard for your code. Whenever you use AI tools like Cursor or Claude Code, i

Neura lets you run long, tool heavy coding sessions without blowing up token costs by compressing old context into references and only resto

Context optimization layer for LLMs. 65-90% token savings with zero quality loss. Drop-in proxy for Claude, GPT, Gemini, and local LLMs (Oll

Claude Code skill — delegate coding tasks to Opencode, supervise via git diff & save 50-90% Claude token usage