Intelligence products that run themselves.

The live momentum index for AI infrastructure.

Open-source AI-infra — agent frameworks, inference engines, vector DBs, gateways — ranked daily by real GitHub velocity, not stars.

Every MCP server, in one living index.

The official Model Context Protocol registry — every server, categorized, freshness-tracked, and searchable. Sourced direct, refreshed every day.

The open-source coding-agent race.

A live leaderboard of open-source coding agents, ranked by shipping velocity — commit cadence, release recency, and trend. Who's actually moving.

What's surging on Hugging Face right now.

The AI models trending on Hugging Face — ranked by trending score, with downloads, likes, and freshness, categorized by task. The pulse of open models.

Every AI-builder index, in one place.

The hub for the whole self-updating fleet — seventeen living indexes of the AI-builder ecosystem, aggregated with live counts and recomputed every day.

Every Claude Code & AI-agent skill, ranked by momentum.

Claude Code skills, subagents, slash-commands, hooks and the collections that hold them — ranked daily by GitHub momentum, not stars.

How do you measure an AI?

LLM & agent evaluation frameworks, benchmark suites, observability and red-teaming harnesses — the eval stack, ranked by momentum.

Run the model on your own hardware.

Tools for running LLMs locally and on-device — inference engines, runners, local chat UIs and quantization, ranked by momentum.

The best prompts are version-controlled.

Prompt-engineering resources — curated collections, leaked system prompts, optimizers, frameworks and guides, ranked by momentum.

The retrieval stack, mapped.

Retrieval-augmented-generation tooling — RAG frameworks, vector databases, embeddings, reranking and ingestion, ranked by momentum.

The generative-imagery stack, developed.

AI image & video generation — diffusion models, ComfyUI nodes, video generators, editing, control and fine-tuning, ranked daily by GitHub momentum.

The voice & speech stack, on air.

Voice & speech AI — text-to-speech, speech recognition, voice cloning & conversion, realtime voice agents and toolkits, ranked daily by GitHub momentum.

The data-centric AI stack, catalogued.

Data-centric AI tooling — data labeling & annotation, synthetic data, curation & quality, augmentation and dataset frameworks, ranked daily by GitHub momentum.

Forge your own model.

LLM fine-tuning & post-training tooling — frameworks, PEFT/LoRA, RLHF/DPO and training data, ranked by momentum.

The layer between your app and every model.

LLM gateways, proxies & routers — the routing layer between apps and model providers: gateways & proxies, model routers, multi-provider SDKs and caching, ranked by momentum.

The stack for building autonomous agents.

AI agent frameworks — the libraries that turn an LLM into a system that can plan, use tools, remember and act: multi-agent, autonomous agents, orchestration, memory and agent SDKs, ranked by momentum.

Turn any document into structured data.

Document-AI tooling — the tools that read a page so your pipeline doesn't have to: OCR engines, PDF extraction, document parsing, layout & table analysis, and vision-language understanding, ranked by momentum.

The production-AI control room.

LLM & AI observability — the tools that watch AI in production: LLM & agent tracing, monitoring & cost analytics, online evaluation, agent observability, and ML / data-drift detection, ranked by momentum.