14,252 mods in this category, of every kind an agent can take. Each one carries what it costs per session, what the scan found, and whether it is the original.
Skill Claude CodeCodex
End-to-end crop disease detection skill supporting classification (which disease), detection (where is the disease), and segmentation (disease region) at three granularity levels, with multiple SOTA models and training strategies built in.
Skill Claude CodeCodex
Production-ready token optimization: reduce costs 40–75% through retrieval pruning, smart caching, and model routing. Use whenever optimizing API costs, latency, or managing long context—especially for RAG pipelines, high-volume systems, multi-turn conversations, or when context exceeds 2K tokens.
Cursor rule Cursor
Cursor rule "ai-usage" from deco-cx/react-tailwind-views, covering ai usage patterns in deco apps, 🏗️ data flow architecture, 🔧 server setup: tool proxying, why proxy ai tools? and implementation pattern.
MCP server Claude CodeCodexCursor +2
A lean, local code-and-artifact knowledge graph that speaks SPARK. Runs locally from the aspark-graph Python package.
MCP server Claude CodeCodexCursor +2
Model Context Protocol server (stdio + remote Streamable HTTP) that lets agents save .md files (with images) straight into md-log. A thin authenticated HTTP client over the hosted md-log service. Runs locally from the md-log-mcp npm package. Needs 2 environment variables to run.
Skill Claude CodeCodex
Use when asked what is new, trending, hot, or worth using in AI right now - which repos, models, papers or launches matter this week, or whether a specific project is gaining traction. Answers from a ranked board with evidence instead of from training data or search-engine marketing.
MCP server Claude CodeCodexCursor +2
A multi-interface (REST and MCP) server for natural language inference. Runs locally from the omni-nli Python package.
Skill Claude CodeCodex
Design feedback mechanisms that help AI systems learn from users - thumbs up/down, preference ranking, corrections, and human-in-the-loop escalation. Use when: RLHF UX, user feedback for AI, thumbs up down design, AI correction flow, human in the loop, feedback signal design, AI improvement loops.
Plugin Claude Code
Bundles 1 skill, 4 commands · 74 tokens together
Simple Local RAG based on memvid.
echoyu1025-a11y/ai-product-skills
Skill Claude CodeCodex
A planning guide for turning an AI product idea into a written product architecture. It organizes the idea around users, situations, capabilities, and data, then explains how information moves through the product.
Plugin Claude Code
Bundles 5 skills · 614 tokens together
Community skills for Claude Code — humanize text, engineer prompts, tune SQL Server queries, write LinkedIn articles, and create Excalidraw diagrams.
Skill Claude CodeCodex
Use when a web app (especially a Fabric App SPA) must read EXISTING Lakehouse data live with no backend — 'query the Lakehouse from the browser', 'live data without a server', 'call Fabric GraphQL from my SPA', or when debugging that path: 403 InsufficientPrivileges on a Fabric GraphQL API, MSAL sign-in returning to a…
Plugin Claude Code
Bundles 1 skill, 1 command, 1 hook · 18 tokens together
Cleans up rough prompts before the assistant acts on them: fixes typos, expands terse asks, repairs broken {{template}} placeholders, and strips filler openers. Includes a /fix-prompt command, an auto-invokable skill, and a non-blocking UserPromptSubmit nudge.
H-Ali13381/recursive-agent-improvement
Skill Claude CodeCodex
Use when an agent repeatedly fails at a narrow measurable subtask, the user wants a trainable specialist tool instead of more prompting, or the task needs an evaluator, reranker, classifier, detector, verifier, or small PyTorch model exposed to an agent.
Agent Claude Code
ML engineering specialist sub-agent. Designs and implements inference pipelines, provider selection, quantization strategies, and hardware abstraction for ONNX models. Benchmark-driven — measures before optimizing.
Plugin Claude Code
A guided data-science lifecycle for Claude Code, with leakage and honesty checks built into every stage.
Plugin Claude Code
Bundles 17 skills · 1,898 tokens together
Skills for the end-to-end ML lifecycle: target review, dataset audits, baselines, experiment planning, training supervision, failure diagnosis, evaluation, explainability, and Go/No-Go gating.
Codeblockz/langchain-community-plugin
Plugin Claude Code
Bundles 3 skills, 5 commands, 2 agents, 1 MCP server · 334 tokens together
Unofficial Claude Code plugin for LangChain and LangGraph. Agents, RAG pipelines, and LCEL chains with best practices.
Liu-MingH/Scientific-research-SKILL
Skill Claude CodeCodex
Structured AI-assisted scientific code generation. 6 safety guards, 8 principles, 11 prompt templates. Grounded in Nature (2026).
XingjianTao/Cursor-Rules-for-PyTorch-DeepLearning-Beginner
Cursor rule Cursor
A Chinese-language set of best practices for PyTorch, a Python framework used to build and train machine-learning models. It describes suggested folders for data, models, source code, experiments, tests, scripts, and configuration files.
Skill Claude CodeCodex
Use this skill when designing how an AI agent should communicate with developers — intent clarification, progress reporting, ambiguity handling, confirmation before irreversible actions, error communication, self-correction, task tracking, and plan mode. Also use when building agent UX or defining agent behavior…
MCP server Claude CodeCodexCursor +2
MCP server "dify-mcp-server" as configured in teddylee777/dify-mcp-server. Runs locally from the dify-mcp-server Python package.
MCP server Claude CodeCodexCursor +2
MCP server that proxies image analysis to OpenRouter vision models. Runs locally from the @forloopcodes/visionmcp npm package.
wang-junjian/ragflow-mcp-server
MCP server Claude CodeCodexCursor +2
RAGFlow MCP Server - A Model Context Protocol server integrating with RAGFlow APIs. Runs locally from the ragflow-mcp-server Python package.
At most 3 mods per repository are shown here, and a mod shipped inside a plugin is left to that plugin's page — the rest are on their repository pages: