Directory
Explore the directory
Tools, libraries, services and references, each reviewed before it goes public. 6,821 resources listed.
Pydantic Deep Agents
vstorm-co.github.ioPydantic Deep Agents is both a self-hosted, open-source terminal AI assistant in the style of Claude Code and the Python framework behind it, so a single function call gives any model a filesystem, shell, planning, memory, sub-agents, sandboxed execution and MCP support.
OCTO Server
github.comOCTO Server is the Go backend at the center of the OCTO platform, exposing REST and WebSocket APIs, orchestrating OpenClaw-powered AI agents called Lobsters as first-class conversation participants, and driving the WuKongIM instant-messaging core for real-time messaging.
Patter SDK
getpatter.comPatter is an open-source SDK that gives an AI agent a phone number, handling the agent loop, speech-to-text, text-to-speech, real-time voice and the telephony carrier in one Python or TypeScript API so a developer can swap any provider in that stack with one line.
Foxel
foxel.ccFoxel is a self-hosted, extensible private cloud storage platform for individuals and teams that unifies local storage, S3, WebDAV, SFTP, FTP and cloud drives such as Google Drive and OneDrive behind one interface, adding AI-powered natural-language search over file contents.
Skills
github.comA collection of agent skills for Kotlin, Jetpack Compose and Android development, packaged as a portable plugin bundle installable in Claude Code, Codex, OpenCode and similar clients.
Ollama-rs
github.comOllama-rs is a Rust client library for the Ollama API, covering text completion including streaming, chat, embeddings, function calling and model management, built by following Ollama's own API documentation.
Awesome MCP Servers
github.comA curated, community list of Model Context Protocol servers, tools, frameworks and clients, translated into five languages, alongside a separately compiled catalog the maintainers describe as covering around six thousand MCP servers found by a web crawler.
Dynamiq
getdynamiq.aiDynamiq is a Python framework for building retrieval-augmented-generation pipelines and LLM agents, with components for code-sandbox tool use, memory-backed chatbots, parallel and sequential multi-agent workflows, and document indexing against a vector database.
OpenOps
openops.comOpenOps is a no-code FinOps automation platform that turns cloud cost-optimization playbooks, such as allocation, anomaly management and safe de-provisioning, into customizable workflows, with a built-in spreadsheet-like database and its own analytics and visualization layer.
Loki Mode
autonomi.devLoki Mode is a free, source-available autonomous coding agent that takes a spec, a GitHub issue, an API document or a one-line brief and builds a deployed application, refusing to call a task done while tests fail or the diff is empty.
numbat
github.comnumbat gives visibility into what AI coding agents do on a developer's machine, watching supported desktop, CLI, IDE and gateway agents through local hooks, logs and session files, and can optionally block a risky action before it runs.
FINN
xilinx.github.ioFINN is an experimental open-source compiler from AMD Research that generates dataflow-style FPGA accelerators for quantized neural networks, aiming for high inference throughput and low latency.
TurboOCR
turboocr.comA GPU document parser that combines OCR, layout analysis, table and formula recognition into one CUDA and TensorRT pipeline, converting pages to Markdown at up to hundreds of images per second.
Agent_Memory_Techniques
diamantai.substack.comAgent Memory Techniques is a set of 30 runnable Jupyter notebooks that teach every major approach to memory for LLM agents, from conversation buffers and vector stores through knowledge graphs and systems such as MemGPT, Mem0, Letta, Zep and Graphiti, plus production patterns.
Ongrid
ongrid.cloudOngrid is a self-hosted operations AI agent that understands an organization's infrastructure, correlates metrics, logs and traces to find a root cause, and can act on it from Slack, Telegram, Lark or DingTalk, with no inbound ports required on the monitored hosts.
MiMo-Audio
xiaomimimo.github.ioMiMo-Audio is an audio language model trained on more than one hundred million hours of audio that shows few-shot generalization across audio tasks it was not explicitly trained for, such as voice conversion, style transfer and speech editing.
ClawVault
github.comClawVault is a security layer for OpenClaw-style AI agents that logs and can block what the agent touches, letting a user lock private files, API keys and credentials into a monitored vault and get an instant-messaging alert whenever the agent reaches for them.
skill-up
alibaba.github.ioskill-up is a CLI that evaluates Agent Skills, agents and their workspaces by running declarative test cases against an agent engine, grading the response and any workspace changes, and producing reports locally or in CI.
A curated marketplace and list of plugins and skills for OpenAI Codex and ChatGPT, covering both official and community-built integrations.
Dive into Pi
build-your-own-pi-cn.enochzhang.chatgpt.siteA Chinese-language course that has the reader build a Pi-style coding agent from scratch across 15 runnable checkpoints, each pairing course text with a real commit, a focused test and a failure experiment.
Dao Code
github.comDao Code is an open-source terminal coding agent built around DeepSeek V4 that reads and writes code, runs commands and fixes bugs behind an approval gate, engineered around DeepSeek's cheap prompt caching to keep memory and self-correction low-cost.
CCCC
cccc.shCCCC coordinates several coding agents, such as Claude Code, Codex and ChatGPT Web, as one persistent team by recording every message in an append-only ledger, tracking delivery and read status, and exposing that state to a web UI, CLI and MCP tools.
Claude Code Chat
github.comClaude Code Chat is a VS Code and Cursor extension that replaces Claude Code's command-line interface with a graphical chat panel, adding checkpoint restore, an inline diff viewer, and a marketplace for browsing and installing MCP servers, skills and plugins.
SKRL
skrl.readthedocs.ioskrl is an open-source, modular reinforcement-learning library implemented in PyTorch, JAX and NVIDIA Warp, designed for readable, transparent algorithms, with support for Gymnasium, Gym, PettingZoo, Isaac Lab and MuJoCo Playground environments.