工具说明为英文。
令牌效率
降低令牌成本并压缩上下文的工具。
用途
活跃度
排序方式
43
工具排名
按实测增长排序,并在不同来源间归一化。 原始值保留各自的数据窗口;计算变化需要 7 天内至少两次测量。
- 1活跃
llmtrim
MCPMCP server and proxy that compresses LLM prompts, tool output, and replies to cut token cost.
mcp实测增长打开来源 ↗
安装
claude mcp add llmtrim -- npx @llmtrim/cli222GitHub 星标+6 (+2.8 %) - 2活跃
rtk
CLICLI proxy that reduces LLM token consumption by 60-90% on common dev commands. Single Rust binary, zero dependencies
github实测增长打开来源 ↗
78 349GitHub 星标+703 (+0.91 %) - 3活跃
headroom
MCPCompress tool outputs, logs, files, and RAG chunks before they reach the LLM. 20% fewer tokens for coding agents, 60-95% fewer tokens for JSON, same answers. Library, proxy, MCP server.
github实测增长打开来源 ↗
68 629GitHub 星标+784 (+1.2 %) - 4活跃
catalyst
技能Token-efficient Claude Code workspace with parallel agents and persistent memory. Research → Plan → Implement → Validate workflow.
github实测增长打开来源 ↗
安装
/plugin marketplace add coalesce-labs/catalyst20GitHub 星标稳定 - 5活跃
Lossless context compression: 2-8x fewer tokens, byte-exact recovery, search inside payloads
mcp实测增长打开来源 ↗
安装
claude mcp add densely -- uvx densely6GitHub 星标稳定 - 6活跃
basecamp
技能A production-ready Claude Code setup. Global CLAUDE.md, coding rules, per-project memory, and token optimization.
github实测增长打开来源 ↗
安装
/plugin marketplace add janmaaarc/basecamp7GitHub 星标稳定 - 7活跃
voly
智能体Control plane for AI coding agents: route tasks, reduce token spend, run multi-agent workflows, fallback executors, and track cost per task.
github实测增长打开来源 ↗
16GitHub 星标稳定 - 8活跃
mcp-recall compresses MCP tool outputs (94 KB → 3.5 KB · 96%) and stores full results in SQLite for retrieval — up to 30x more tool calls per session for heavy MCP workloads.
github实测增长打开来源 ↗
安装
/plugin marketplace add sakebomb/mcp-recall8GitHub 星标稳定 - 9活跃
Staged context pruning for OpenCode, Oh My Pi, pi, and Claude Code.
github实测增长打开来源 ↗
安装
/plugin marketplace add AshishKumar4/better-compact12GitHub 星标+1 (+9.1 %) - 10活跃
Governance framework for AI coding agents. It runs them through a five-step workflow (plan, build, review, test, ship) where no step counts as done without evidence. Drop-in rules and guardrails for Claude Code, Codex, Cursor, Copilot, and…
github实测增长打开来源 ↗
安装
git clone https://github.com/KbWen/agentic-os ~/.claude/skills/agentic-os155GitHub 星标+8 (+5.4 %) - 11活跃
Persistent memory + security layer for Claude Code (MCP plugin). Agents remember your project across sessions (~87% fewer context tokens), every tool call lands in a tamper-evident HMAC audit chain, and skills are AST-scanned +…
github实测增长打开来源 ↗
8GitHub 星标稳定 - 12活跃
对标 claude-hud 核心功能,控制在 7 行以内更轻量。内置 13 个国产大模型平台用量查询 (智谱 GLM / MiniMax / DeepSeek / Kimi / StepFun / SiliconFlow 等),第三方代理用户也能实时看到额度消耗。
github实测增长打开来源 ↗
安装
/plugin marketplace add zander-zyx/claude-mini-hud9GitHub 星标稳定 - 13活跃
Lossless context compression plugin for Claude Code & OpenCode.
github实测增长打开来源 ↗
安装
/plugin marketplace add aerovato/magic-compact146GitHub 星标+1 (+0.69 %) - 14活跃
Content-aware output compression for AI coding assistants. 36 specialized processors cut CLI output tokens by 60-99% (git, pytest, npm, terraform, kubectl, docker, and more) without losing errors, diffs, or stack traces.
github实测增长打开来源 ↗
安装
/plugin marketplace add ppgranger/token-saver142GitHub 星标+2 (+1.4 %) - 15活跃
Statically audits MCP tool surfaces for token cost, schema quality, and design issues.
mcp实测增长打开来源 ↗
需在 MCP 配置中声明的远程服务器
https://mcplint-web.vercel.app/api/mcp0GitHub 星标稳定 - 16活跃
Verifies token/cost savings claimed by AI context-reduction proxies via MCP tools.
mcp实测增长打开来源 ↗
安装
claude mcp add tokentrust -- uvx tokentrust-cli0GitHub 星标稳定 - 172 122GitHub 星标+108 (+5.4 %)
- 18活跃
Claude Code skill — delegate coding tasks to Mistral Vibe, supervise via git diff & save 50-90% Claude token usage
github实测增长打开来源 ↗
安装
git clone https://github.com/pcx-wave/vibe-skill ~/.claude/skills/vibe-skill117GitHub 星标稳定 - 19活跃
Tinder for your Claude Code skills, subagents and MCP servers. Swipe away what wastes context, scan for prompt injection, get honest token costs. Fixes what it finds and actually deletes. Zero dependencies, Claude Code + Codex.
github实测增长打开来源 ↗
115GitHub 星标+1 (+0.88 %) - 20活跃
Claude Code plugin that tracks token usage, identifies wasted context, and saves 30-50% on API costs. Heatmaps, ROI reports, budget alerts, efficiency scores, git-aware suggestions — all local, zero config.
github实测增长打开来源 ↗
安装
/plugin marketplace add egorfedorov/claude-context-optimizer103GitHub 星标+6 (+6.2 %) - 21活跃
Tokdash
技能Agent Dashboard: Visualization and analytics for Sessions and Quota Usage. Track, analyze, and optimize token usage across providers with heatmaps, cost tracking, token counting and quota resets..
github实测增长打开来源 ↗
安装
/plugin marketplace add JingbiaoMei/Tokdash67GitHub 星标+3 (+4.7 %) - 22活跃
Measure and shrink the token cost of Claude Code skills, tier by tier: the always-on system-prompt tax, the per-invocation body, and on-demand references, each priced at the rate it is actually billed.
github实测增长打开来源 ↗
1GitHub 星标稳定 - 23活跃
toonify-mcp
MCPContext compression plugin for Claude Code. Automatically trims large tool output—JSON, YAML, stack traces, and logs—before it enters the context window.
github实测增长打开来源 ↗
64GitHub 星标稳定 - 24活跃
skillreaper
MCPCut context bloat in your AI-agent stack: find and safely prune unused skills, MCP servers and subagents from real transcript evidence
github实测增长打开来源 ↗
55GitHub 星标+1 (+1.9 %) - 25活跃
hush
技能🤫 Token-lean sessions at the harness level. An easy to grasp output style, output-shrinking hooks, and log compression cut both input and output tokens.
github实测增长打开来源 ↗
安装
/plugin marketplace add V-Songbird/hush47GitHub 星标+4 (+9.3 %)