28 boosters for "cost" — open source, verified from GitHub, ready to install
An intelligent API performance governor that autonomously optimizes system execution while preventing cost overruns and security breaches through strict guardrails. Ideal for developers managing cloud APIs and ML systems where runaway costs are a critical concern.
Infrastructure Maintainer is an expert agent that helps teams design, monitor, and optimize cloud infrastructure for reliability, performance, and cost efficiency. DevOps engineers, platform teams, and SREs can use it to troubleshoot systems, plan scaling strategies, and maintain high availability.
Train object detection, image classification, and SAM/SAM2 segmentation models on managed cloud GPUs. No local GPU setup required—results are automatically saved to the Hugging Face Hub. Use this skill when users want to: Helper scripts use PEP 723 inline dependencies. Run them with :
Train language models using TRL (Transformer Reinforcement Learning) on fully managed Hugging Face infrastructure. No local GPU setup required—models train on cloud GPUs and results are automatically saved to the Hugging Face Hub. Use this skill when users want to: Use Unsloth () instead of standard
Run any workload on fully managed Hugging Face infrastructure. No local setup required—jobs run on cloud CPUs, GPUs, or TPUs and can persist results to the Hugging Face Hub. Use this skill when users want to: When assisting with jobs:
This skill enables users to run Python workloads, Docker jobs, and GPU-intensive tasks on Hugging Face's managed infrastructure without local setup. It's valuable for ML engineers, data scientists, and developers needing cloud compute for training, inference, and batch processing.
A skill for fine-tuning and training language models on Hugging Face's cloud GPU infrastructure using TRL, supporting SFT, DPO, GRPO methods and GGUF conversion for local deployment. Developers and ML engineers working with cloud-based model training benefit from this comprehensive guidance.
Yielding Bear provides a single unified API that routes every LLM request to the cheapest capable model across 16+ providers — saving 60-80% vs calling OpenAI, Anthropic, or Google directly. 1. Get an API key at https://yieldingbear.com/api 2. Set environment variable:
"name": "nano-banana-2-skill-marketplace", "url": "https://github.com/kingbootoshi" "description": "AI image generation CLI powered by Gemini 3.1 Flash (default) and Gemini 3 Pro. Multi-resolution, aspect ratios, cost tracking, green screen transparency, reference images, style transfer.",
"name": "claudetop", "description": "htop for your Claude Code sessions — real-time cost tracking, cache efficiency, model comparison, smart alerts, web dashboard, and session analytics", "name": "Lior Weinstein",
This skill should be used when users want to train or fine-tune language models using TRL (Transformer Reinforcement Learning) on Hugging Face Jobs infrastructure. Covers SFT, DPO, GRPO and reward modeling training methods, plus GGUF conversion for local deployment. Includes guidance on the TRL Jobs package, UV scripts with PEP 723 format, dataset preparation and validation, hardware selection, cost estimation, Trackio monitoring, Hub authentication, and model persistence. Should be invoked for tasks involving cloud GPU training, GGUF conversion, or when users mention training on Hugging Face Jobs without local GPU setup.
You are a Pinocchio framework specialist focused on extreme CU optimization and minimal binary size for Solana programs. You write zero-copy, hand-optimized code that achieves 80-95% CU savings vs Anchor. Use Mollusk for fast, isolated instruction testing: use mollusk_svm::Mollusk;
"name": "harness-mcp-v2", "description": "Give AI agents full access to the Harness.io platform — manage pipelines, deployments, cloud costs, chaos engineering, feature flags, SEI, and 125+ resource types through 11 MCP tools", "main": "build/index.js",
"name": "claude-context-optimizer", "description": "ContextShield, CLAUDE.md analyzer, interactive dashboard, confidence learning, and token optimization for Claude Code", "name": "Egor Fedorov"
CU optimization specialist using Pinocchio framework. Use for performance-critical programs requiring 80-95% CU reduction vs Anchor. Specializes in zero-copy access, manual validation, and minimal binary size.\n\nUse when: CU limits are being hit, transaction costs are significant at scale, binary size must be minimized, or maximum throughput is required.
"description": "WozCode enhanced coding tools — smart search, batch editing, SQL introspection, and cost-optimized subagent delegation", "version": "0.3.18", "homepage": "https://withwoz.com"
HiveMind-Actions is a GitHub Actions prompt that enables multi-agent AI code reviews by analyzing diffs against project rules and outputting structured JSON feedback. It benefits developers who want automated, rule-based code quality checks integrated directly into their CI/CD workflows.
HiveMind-Actions is a serverless swarm AI system for GitHub that automates issue analysis and creates execution plans by coordinating multiple AI agents. It benefits development teams seeking to reduce infrastructure overhead while automating code review, issue triage, and implementation planning.
Markdown Agents is a framework for defining AI agents using simple markdown files with YAML frontmatter, enabling version-controlled, portable agent configuration across Claude Code and Desktop. Developers benefit from reduced cognitive load, git-native memory tracking, and predictive AI collaboration without session amnesia.
"name": "codesession-marketplace", "name": "Brian Mwirigi", "email": "brianmwirigi@users.noreply.github.com"
Finance Tracker is an AI agent that manages studio budgets, optimizes costs, forecasts revenue, and analyzes financial performance metrics to maximize resource ROI. It benefits finance teams, studio managers, and project leads who need data-driven financial insights.
"description": "Save tokens and cut inference costs by routing compute through MVM nodes on the cheapest available energy.", "url": "https://github.com/nhevers" "repository": "https://github.com/nhevers/mica-plugin",
Intelli-router automatically triages user messages by complexity and routes them to the most cost-effective AI model (local Ollama, Claude Sonnet, Codex, or Claude Opus). Developers building multi-model AI systems benefit from reduced costs and optimized latency.
Heuristic scoring (no AI key configured).
