prune: sweep AI and Agents to the shortlist cap

Per ADR-0001, restructure first: split Pre-trained Models and Inference
into Pre-trained Models, LLM Inference and Serving, and LLM Gateways;
mint Vendor Agent SDKs, Personal Assistants, Prompt Optimization, Image
and Video Generation, and Fine-tuning subcategories. Re-home
openai-agents (Vendor Agent SDKs), hermes-agent (Personal Assistants,
description reworded to personal assistant), dspy (Prompt Optimization),
diffusers (Image and Video Generation), unsloth (Fine-tuning), and
graphify (Data Visualization > Specialized). Then cap.

Removed (downloads are PyPI last-month via pypistats, 2026-08-14/15):

- nuwa-skill: persona prompts, not engineering
- crewai (19.4M/month): buzz peaked; no longer named unprompted
- autogen (1.1M/month): fork war with ag2 split its community
- ag2 (441K/month): other half of the same fork war
- smolagents (651K/month): Hugging Face ecosystem niche
- TradingAgents (11.9K/month): vertical trading app, not general
  orchestration
- bub (5.6K/month): no adoption
- bindu (210/month): no adoption
- livetalking: no PyPI presence, no adoption evidence
- bernstein: no PyPI presence, no adoption evidence
- promptise: no PyPI presence, no adoption evidence
- OpenChronicle: no PyPI presence, no adoption evidence
- outlines (2.5M/month): same job as instructor, which holds the slot
- entroly (3.1K/month): no adoption
- lumen (1.8K/month): no adoption
- liter-llm (3.5K/month): xberg-io coordinated self-promotion plant
- SenseVoice: whisper wins the use case
- voxcpm (101K/month): no track record

Co-Authored-By: Claude <noreply@anthropic.com>

README.md

@@ -134,50 +134,38 @@ _Libraries for building AI applications, LLM integrations, and autonomous agents
 
 - Agent Skills
   - [django-ai-plugins](https://github.com/vintasoftware/django-ai-plugins) - Django backend agent skills for Django, DRF, Celery, and Django-specific code review.
-  - [graphify](https://github.com/safishamsi/graphify) - Turn any folder of code, SQL schemas, docs, papers, images, or videos into a queryable knowledge graph.
-  - [nuwa-skill](https://github.com/alchaincyf/nuwa-skill/blob/main/README_EN.md) - Nuwa distills the thinking of anyone — let Musk, Naval, Munger, and Feynman work for you.
   - [sentry-skills](https://github.com/getsentry/skills) - Python-focused engineering skills for code review, debugging, and backend workflows.
   - [trailofbits-skills](https://github.com/trailofbits/skills) - Python-friendly security skills for auditing, testing, and safer backend development.
 - Orchestration
-  - [ag2](https://github.com/ag2ai/ag2) - An open-source AgentOS for multi-agent orchestration and building agentic AI systems.
-  - [autogen](https://github.com/microsoft/autogen) - A programming framework for building agentic AI applications.
-  - [bernstein](https://github.com/sipyourdrink-ltd/bernstein) - A deterministic Python orchestrator for CLI coding agents (Claude Code, Codex, Gemini CLI, and 40+ more) with parallel git worktrees and an HMAC-signed audit chain.
-  - [bindu](https://github.com/getbindu/Bindu) - A framework that wraps any agent handler with DID-based cryptographic identity, A2A JSON-RPC over HTTP, OAuth2 auth, x402 (USDC) payments, and a built-in operator inbox.
-  - [bub](https://github.com/bubbuild/bub) - A lightweight, hook-first Python framework for channel-native agents that live alongside people.
-  - [crewai](https://github.com/crewAIInc/crewAI) - A framework for orchestrating role-playing autonomous AI agents for collaborative task solving.
-  - [dspy](https://github.com/stanfordnlp/dspy) - A framework for programming, not prompting, language models.
-  - [hermes-agent](https://github.com/nousresearch/hermes-agent) - An adaptive AI agent framework that grows with you.
   - [langchain](https://github.com/langchain-ai/langchain) - Building applications with LLMs through composability.
   - [langgraph](https://github.com/langchain-ai/langgraph) - Low-level orchestration framework for building stateful, long-running LLM agents.
-  - [livetalking](https://github.com/lipku/livetalking) - A framework for real time interactive streaming digital human.
-  - [openai-agents](https://github.com/openai/openai-agents-python) - OpenAI's framework for building and managing AI agents.
-  - [OpenChronicle](https://github.com/Einsia/OpenChronicle) - Open-source, local-first memory for any tool-capable LLM agent.
-  - [promptise](https://github.com/promptise-com/foundry) - A framework for building end-to-end production-ready agentic systems, scalable & secure MCP's and autonomous agents.
   - [pydantic-ai](https://github.com/pydantic/pydantic-ai) - A Python agent framework for building generative AI applications with structured schemas.
-  - [smolagents](https://github.com/huggingface/smolagents) - A lightweight library to build agents and multi-agent workflows with minimal code.
-  - [TradingAgents](https://github.com/TauricResearch/TradingAgents) - A multi-agents LLM financial trading framework.
+- Vendor Agent SDKs
+  - [openai-agents](https://github.com/openai/openai-agents-python) - OpenAI's framework for building and managing AI agents.
+- Personal Assistants
+  - [hermes-agent](https://github.com/nousresearch/hermes-agent) - An adaptive personal AI assistant that grows with you.
+- Prompt Optimization
+  - [dspy](https://github.com/stanfordnlp/dspy) - A framework for programming, not prompting, language models.
 - Data Layer
-  - [entroly](https://github.com/juyterman1000/entroly) - An auditable context control plane that optimizes prompt context, stabilizes cache prefixes, and verifies answers locally with WITNESS.
   - [instructor](https://github.com/567-labs/instructor) - A library for extracting structured data from LLMs, powered by Pydantic.
   - [llama-index](https://github.com/run-llama/llama_index) - A data framework for your LLM application.
-  - [lumen](https://github.com/holoviz/lumen) - An agent framework for conversational analysis of SQL, dataframe and multidimensional array data.
   - [mem0](https://github.com/mem0ai/mem0) - An intelligent memory layer for AI agents enabling personalized interactions.
-  - [outlines](https://github.com/dottxt-ai/outlines) - Structured text generation for LLMs with JSON schema, regex, and grammar-constrained decoding.
-- Pre-trained Models and Inference
-  - [diffusers](https://github.com/huggingface/diffusers) - A library that provides pre-trained diffusion models for generating and editing images, audio, and video.
-  - [LiteLLM](https://github.com/BerriAI/litellm) - Call 100+ LLMs using OpenAI format.
-  - [liter-llm](https://github.com/xberg-io/liter-llm) - A universal LLM API client with a Rust core, calling 142+ providers through a single OpenAI-compatible interface, with a built-in proxy and MCP server.
+- Pre-trained Models
+  - [transformers](https://github.com/huggingface/transformers) - A framework that lets you easily use pre-trained transformer models for NLP, vision, and audio tasks.
+- LLM Inference and Serving
   - [mlx-lm](https://github.com/ml-explore/mlx-lm) - Run and fine-tune large language models on Apple Silicon with MLX.
   - [sglang](https://github.com/sgl-project/sglang) - A high-performance serving framework for large language models and multimodal models.
-  - [transformers](https://github.com/huggingface/transformers) - A framework that lets you easily use pre-trained transformer models for NLP, vision, and audio tasks.
-  - [unsloth](https://github.com/unslothai/unsloth) - A library for faster LLM fine-tuning and training with reduced memory usage.
   - [vllm](https://github.com/vllm-project/vllm) - A high-throughput and memory-efficient inference and serving engine for LLMs.
+- LLM Gateways
+  - [LiteLLM](https://github.com/BerriAI/litellm) - Call 100+ LLMs using OpenAI format.
+- Image and Video Generation
+  - [diffusers](https://github.com/huggingface/diffusers) - A library that provides pre-trained diffusion models for generating and editing images, audio, and video.
+- Fine-tuning
+  - [unsloth](https://github.com/unslothai/unsloth) - A library for faster LLM fine-tuning and training with reduced memory usage.
 - Speech
   - [funasr](https://github.com/modelscope/FunASR) - Industrial-grade speech recognition toolkit with 170x realtime speed, 50+ languages, speaker diarization, and emotion detection.
   - [openai-whisper](https://github.com/openai/whisper) - A general-purpose automatic speech recognition model trained on 680k hours of multilingual and multitask supervised data.
-  - [SenseVoice](https://github.com/FunAudioLLM/SenseVoice) - A multilingual speech understanding model for speech recognition, language identification, emotion recognition, and audio event detection.
   - [vibevoice](https://github.com/microsoft/VibeVoice) - A family of open-source voice AI models from Microsoft for text-to-speech and long-form speech recognition.
-  - [voxcpm](https://github.com/OpenBMB/VoxCPM) - A tokenizer-free text-to-speech foundation model for multilingual speech generation and voice cloning.
 
 ### Deep Learning
 
@@ -544,6 +532,7 @@ _Libraries for visualizing data. Also see [awesome-javascript](https://github.co
   - [vispy](https://github.com/vispy/vispy) - High-performance scientific visualization based on OpenGL.
 - Specialized
   - [cartopy](https://github.com/SciTools/cartopy) - A cartographic python library with matplotlib support.
+  - [graphify](https://github.com/safishamsi/graphify) - Turn any folder of code, SQL schemas, docs, papers, images, or videos into a queryable knowledge graph.
   - [pygraphviz](https://github.com/pygraphviz/pygraphviz/) - Python interface to [Graphviz](https://www.graphviz.org/).
 - Dashboards and Apps
   - [gradio](https://github.com/gradio-app/gradio) - Build and share machine learning apps, all in Python.