AI Train2025
Trains multi-step LLM agents with reinforcement learning (GRPO) on your own tasks, wrapping existing agent code behind an OpenAI-compatible client. Its RULER mode scores trajectories with an LLM judge, so there's no reward function to hand-write.
AI Agent2025
Provides 7×24 automated customer service for the Xianyu marketplace with multi-expert routing, context-aware dialogue, and a laddered bargaining system. Built in Python and designed to run against an LLM provider with browser-cookie integration for web interactions.
AI Infra2025
Transforms unstructured financial content—papers, news, blogs, and filings—into a queryable semantic knowledge graph for retrieval-augmented research. Combines domain-tuned LLMs, embedding-based search, and modular ingestion pipelines; aimed at quant research teams and institutional workflows.
MCP Client2025
Bridges LLM-driven AI assistants to the Unity Editor so models can create scenes, edit C# scripts, manage assets, run tests and automate game-dev workflows. Exposes 47 focused MCP tool entrypoints, supports many MCP clients, and is MIT-licensed for local use.
AI Agent2025
Twelve engineering principles for building production-grade LLM agents, modeled on the 12-Factor App. Argues the best agents are mostly deterministic software with a few well-placed LLM calls, not a prompt-and-tools loop.
AI Client2025
Runs open-source LLMs and multimodal models entirely on mobile devices for offline, private inference. Offers Agent Skills, Thinking Mode, Ask Image, audio scribe, model management and benchmarks, with Gemma 4 and Hugging Face integration.
AI Agent2025
Builds a table-of-contents tree index over long PDFs and uses LLM tree search to fetch relevant sections — no embeddings, chunking, or vector database. Hits 98.7% on FinanceBench, for financial, legal, and technical docs where relevance needs reasoning.
AI Agent2025
Provides ready-to-use sample agents for Google’s Agent Development Kit across Python, TypeScript, Go, Java, Kotlin, and Android, from simple assistants to multi-agent workflows.
AI Coding2025
Combines static code analysis with LLM reasoning to produce interactive architecture diagrams, component-level documentation, and navigable outputs for IDEs, CI, and docs. Emits Mermaid diagrams and incremental updates with CLI and editor integrations.
AI Dataset2025
Provides 5 million instruction–response pairs for supervised fine-tuning of code LLMs, with inputs, outputs, unit tests, and automated LLM judgments. Uses hybrid automated/synthetic generation and is released under CC BY 4.0 for large-scale SFT workflows.
AI Others2025
Argues AI has entered its 'second half': a working recipe (language pre-training priors + scale + reasoning) now generalizes RL across tasks, so the bottleneck shifts from inventing methods to defining problems and rethinking evaluation.
AI Agent2025
Lets LLM agents drive real Android and iOS devices from natural-language commands by turning each screen's accessibility tree into structured text the model reads directly, not just screenshots. LLM-agnostic; runs via CLI, Python, or Docker.