Havuz

Cost-efficient coding-agent runtime: prefix-aware prompt-cache routing, poisoning-safe semantic cache, budgeted context compression, request

50x your usage with cheap-ag. Opus/Astra plans, a dirt-cheap model writes every line, and tests decide what ships. Your $20 plan stops hitti

Cost control for long Claude Code sessions on Windows, macOS and Linux: tiered context ceilings, a blocking handoff gate, and effort gears.

Claude Code plugin: see where your tokens go, audit your setup against the docs, and tune settings with an approved diff

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

Experience-Driven Black-Box Cost Control for Coding Agents

麦麦省钱助手 —— 基于麦当劳 MCP 的点餐成本优化 Skill。算出当前门店最省钱的点餐方案，并主动提示凑单陷阱。

🍟 麦麦省钱助手 —— 麦当劳中国官方 MCP 的省钱点餐 Agent Skill：官方 calculate-price 实时核价 + 券组合优化 + 沉默成本体检与多单用券排程，官方券/套餐/团购券/积分四渠道同口径算终局价。免登录在线试算。

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

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

The cheapest way to run GitHub Actions: AWS spot machines in your own account at AWS's price, nothing added ($0.0008 a minute for 2 vCPU). O

Know what your stack will cost before the bill arrives. Free agent skill for Claude Code, Cursor and Codex: prices every service at your usa

Agent Smith, Codex görevleri için yerel model yönlendirmesi, macOS kullanım izleme arayüzü ve tekrarlanabilir benchmark araçları sunuyor.

Neden yükseliyor · Altı günde 194 yıldız alması dikkat çekici; yerel model routing, görünür kullanım takibi ve benchmark sonuçlarını tek projede birleştirmesi ilgi çekmiş olabilir. Ancak bu erken ilgi benimseme kanıtı değil: son 28 günde 5 commit ve 1 farklı yazar var.

Spend fewer tokens on the same work. A local CLI that finds wasted tokens in Claude Code, Codex, opencode and your own API calls.

Budget caps, runaway-loop detection and circuit breakers for AI agents. Zero dependencies, framework agnostic.

Distil an expensive LLM API call on a narrow task into a small local model: capture traffic, curate, LoRA fine-tune with MLX on Apple Silico

Evidence-backed attack paths, drift, and cost auditing for Hetzner Cloud.

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

Make Codex and Claude Code subscriptions go further: a local model router that gives routine agent work to economical models and keeps your

Model router for AI coding agents: routes routine work to cheaper models with routing rules, 8 subagents, hooks and a CLI runner

The ops layer for running Claude Code and Codex all day on a Mac: subscription-first routing with overflow to GLM/OpenRouter/Bedrock, pausel

A repo investigator for Claude Code and Codex. Offload codebase investigation and get up to 2.7× as much work from the same limits.

Sub-millisecond query routing & cost-aware execution for GraphRAG architectures. Adaptive-AtomRAG is a highly optimized conditional routing

Local-first control center for AI-assisted development — observe usage and cost, compare models, code with OpenCode, and understand capacity

Per-turn model routing for the Pi coding agent: a small judge picks the cheap or the strong tier for each message, with multi-model failover

Local-first CLI that reads your Spec-Driven Development artifacts and recommends the best value-for-money AI model to implement them. Powere

Local Anthropic-compatible AI gateway with cheap-first routing, quality gates, safe model handoffs, and an embedded cost dashboard.

A token-spend profiler and cost-regression gate for AI agents.

SKILL: Claude FABLE 5 spanking Claude OPUS 4.8. Orchestrator-worker cost model for Claude Code: the expensive model reads and judges, Opus 4

Open-source Codex plugin for Sol-governed multi-agent coding, bounded AI code review, model routing, and LLM cost optimization

🦴 paleo — token-saving skills for LLM agents: compress output, trim context, cap budget (Claude Code / Codex / Gemini / Hermes)

Cost-optimised agent router for Claude Code: sends every sub-task to the cheapest model that can do it, and escalates only on verified failu

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

Replay real agent traces through cheaper models to prove which swaps are safe.

Unofficial source-linked AWS best practices for every AWS service, organized by Well-Architected pillar for Claude Code, OpenAI Codex, and A

Slash API Batch: Cut AI Costs by 50% in 2026

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

Fugusashi — Federated Learning of LLM Routing with Human-Interpretable Decisions. Open-source alternative to Sakana AI Fugu.

Intelligent multi-LLM router with task-aware routing strategies, cost optimization, and production safety controls — drop-in OpenAI-compatib

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

Free, local, open-source LLM cost analyzer - see where your LLM bill leaks, on your machine.

A prompt-aware LLM router that predicts which models can complete each request, then selects the cheapest capable one: 53.2% lower cost and

Open-source LLM control plane: observe every call, recommend savings, then apply only human-approved, reversible routing.

Local streaming reverse proxy between AI coding agents (Claude Code, Cursor, Codex) and model APIs (Anthropic, OpenAI, Gemini, MiniMax). Met

Long-term memory for AI agents with no LLM call: an MCP server and Claude Code plugin, plus local-first context compression for LLM apps.

Beastmode: MofA (Mixture of Agents) orchestration framework for Hermes/OpenClaw/Codex with MemroOS-style context continuity.

Scan your Claude Code logs and find the habits silently burning your tokens. No LLM calls, no deps, offline.

Compile an AI agent's repeated workflows into deterministic, auditable routines that replay for free, with a fallback to the agent.

Call the Google Antigravity CLI (agy) headlessly from any non-TTY context — Windows ConPTY + POSIX pty + MCP server. Fixes the empty-output

Automatic model+effort routing for Claude Code: tiered subagents, escalation ladder, independent verifier