Packages

PUAClaw is a documentation-first framework for classifying and organizing prompt-based psychological persuasion / unconventional rhetoric techniques used with LLMs. The repository primarily provides taxonomies, technique categories, templates/“prompt” wording examples, and related documentation (e.g., glossary, FAQ, technique index), rather than a software service or API.

Lists third-party LLM inference providers (including model owners and inference hosting platforms) that advertise permanent free tiers, along with example model names and rough rate-limit/free-tier notes and links to each provider’s API key page. It does not itself provide an API; it is a curated catalog.

这是一个中文学习教程仓库，面向 Claude Code（Anthropic 官方编程 CLI）与 OpenClaw（开源 AI 助手/多代理框架）。内容主要是安装、配置、使用、MCP 集成、Hooks/Skills/Plugins、Agent 编排、多平台消息接入、安全与部署等实践文档与示例。

awesome-hacking is a curated list (an “awesome” repository) of hacking/security-related tools and resources. It is not, by itself, an API/service; it primarily provides links and pointers to third-party tools.

A learning repository (“AWS zero to hero”) outlining a 30-day, day-by-day curriculum for AWS DevOps skills. The README describes topics/projects across IAM, EC2/VPC networking, S3, CLI, CloudFormation, CI/CD (CodeCommit/CodePipeline/CodeBuild/CodeDeploy), monitoring (CloudWatch), serverless (Lambda/EventBridge), CDN (CloudFront), containers (ECR/ECS/EKS), secrets (AWS Secrets Manager), infrastructure automation (Terraform), auditing/compliance (CloudTrail/Config), load balancing, and interview preparation. It appears to be educational content rather than a software API/service.

MiniMind is an open-source, end-to-end small LLM training and inference project (PyTorch-focused) covering model architecture (Dense + MoE), tokenizer training, data pipelines, pretraining, SFT, LoRA, and preference/RLHF-style training (e.g., DPO and other variants mentioned), plus a minimal OpenAI-compatible API server and a Streamlit web UI for chat/tool-calling style interactions.

LLaMA Factory (llamafactory) is a Python framework/CLI/UI for training and fine-tuning a wide range of LLMs and multimodal models using many supervised and RL-style training approaches, with support for efficient methods (e.g., LoRA/QLoRA and quantization) and multiple inference backends including an OpenAI-style API via vLLM/SGLang.

OpenClaw Dashboard is a local, DB-less web UI (Next.js) that reads OpenClaw configuration and local session files from ~/.openclaw/openclaw.json (or OPENCLAW_HOME) to display bot/agent status, model details, session lists and token usage trends, gateway health, connectivity tests, alerts, and an i18n/light-dark themed pixel-art “office” visualization.

CC Switch is a cross-platform desktop (Tauri/Rust + TypeScript UI) app that manages and switches configuration for multiple AI coding/assistant CLI tools (Claude Code, Codex, Gemini CLI, OpenCode, OpenClaw). It provides provider presets, unified MCP and Skills management with sync across apps/devices, a proxy/failover layer, and local persistence via SQLite.

exo is a local-first system for running LLM inference across multiple devices by automatically discovering peers and distributing model execution (tensor/pipeline parallelism) over the network, with an optional built-in dashboard and API compatible with common chat/response endpoints. On macOS it also describes RDMA-over-Thunderbolt support for reduced inter-device latency.

U-Claw is a tutorial + source code to package the OpenClaw AI assistant into a portable USB-style application. It provides scripts and a file skeleton to download dependencies (Node.js/Electron and OpenClaw runtime), configure AI models and chat/bot platforms (e.g., QQ/Feishu/Telegram/WhatsApp/Discord/WeChat plugins), and run the assistant via OS-specific launchers (Mac .command, Windows .bat) including a bootable USB option (Ventoy + Ubuntu) and a separate Electron desktop app.

SiYuan is a privacy-first, self-hosted personal knowledge management system for creating and managing knowledge using block-level references and a Markdown/WYSIWYG editor. It supports exporting content and includes features such as search, OCR, spaced repetition, and AI writing/Q&A via external APIs (e.g., OpenAI).

Apache Airflow is an open-source platform to programmatically author, schedule, and monitor workflows (DAGs). It executes tasks via a scheduler and workers, provides a UI for visualization/monitoring, and supports defining pipelines as code.

A GitHub repository presented as an “interview preparation book”/knowledge collection for AIGC/LLMs/AI agents and related algorithm/development topics, organized as a set of markdown sections/directories covering fundamentals and interview question practice; also promotes a community for career/job support.

awesome-ai-agent-builder is a curated “awesome list” repository that links to frameworks and platforms for building AI agents (code-centric, low-code, and no-code). It does not provide a single runnable API/library itself.

LLM/RL/VLM 算法与训练方法的可视化学习资料（主要为大量架构图、SVG/PDF/图片与目录化讲解），用于帮助理解大模型结构、解码、SFT、DPO/RLHF、以及强化学习理论与方法。

This repository is a curated documentation hub of ~74 concrete “OpenClaw AI agent” use cases (skills) sourced from Moltbook, organized by category (everyday life, memory, automation, data analysis, security monitoring, tool development). It primarily provides markdown-based setup prompts, required skills, and success metrics rather than an executable API/SDK.

awesome-mlops appears to be a curated directory (“awesome list”) of MLOps tools grouped by category (e.g., AutoML, CI/CD for ML, monitoring, data cataloging, drift detection). It does not itself provide an API or executable service; it links out to other projects.

awesome-python is a curated, opinionated “awesome list” of Python frameworks, libraries, tools, and resources organized by category (web, ML, data, dev tools, etc.). It is not itself an API service or SDK.

ml-road is a GitHub repository that curates educational resources (courses, books, and links) related to machine learning and agentic AI. It primarily serves as a reading/training roadmap rather than providing an executable ML library or an API service.