# Loong

> Public documentation for Loong, the Rust foundation for vertical AI agents.

## Docs

- [Architecture](https://docs.loongclaw.ai/build-on-loong/architecture.md): The public architecture contract for contributors and integrators.
- [Contributing](https://docs.loongclaw.ai/build-on-loong/contributing.md): Where contributors should start in the public repository.
- [Contribution Areas](https://docs.loongclaw.ai/build-on-loong/contribution-areas.md): Where different kinds of contributors can create real leverage in Loong.
- [Contribution Workflow](https://docs.loongclaw.ai/build-on-loong/contribution-workflow.md): The public contribution bar, validation flow, and standard PR path.
- [Docs Workflow](https://docs.loongclaw.ai/build-on-loong/docs-workflow.md): How Loong's public docs are split, validated, and deployed from this repository.
- [Build On Loong](https://docs.loongclaw.ai/build-on-loong/overview.md): Understand the boundaries you should preserve when extending the system.
- [Doctor And Health](https://docs.loongclaw.ai/get-started/doctor-and-health.md): Diagnose and repair a broken local setup.
- [First Run](https://docs.loongclaw.ai/get-started/first-run.md): Reach a first useful answer without learning the full runtime surface.
- [Installation](https://docs.loongclaw.ai/get-started/installation.md): Install Loong from release artifacts or source.
- [Get Started](https://docs.loongclaw.ai/get-started/overview.md): The shortest path from install to first useful answer.
- [Loong Documentation](https://docs.loongclaw.ai/index.md): Public docs for Loong, the Rust base for vertical AI agents.
- [Documentation Policy](https://docs.loongclaw.ai/reference/documentation-policy.md): How the public docs surface is scoped, written, and deployed.
- [Reference](https://docs.loongclaw.ai/reference/overview.md): Public reference material that remains part of the OSS repository.
- [Releases](https://docs.loongclaw.ai/reference/releases.md): Public release notes and announcements.
- [Roadmap And Product](https://docs.loongclaw.ai/reference/roadmap-and-product.md): Public product direction, first-run contract, and roadmap references.
- [Security And Reliability](https://docs.loongclaw.ai/reference/security-and-reliability.md): Public invariants and disclosure references.
- [Why Loong](https://docs.loongclaw.ai/reference/why-loong.md): Why Loong exists, what stance it takes, and what kind of runtime it is trying to become.
- [Browser And Web Boundaries](https://docs.loongclaw.ai/use-loong/browser-and-web-boundaries.md): Understand private-host, domain, and browser runtime gates without guessing.
- [BytePlus Coding Plus Telegram](https://docs.loongclaw.ai/use-loong/byteplus-coding-telegram-playbook.md): Keep BytePlus coding traffic explicit while Telegram owns the shipped reply-loop surface.
- [DingTalk](https://docs.loongclaw.ai/use-loong/channel-guides/dingtalk.md): Configure the DingTalk channel surface in Loong.
- [Discord](https://docs.loongclaw.ai/use-loong/channel-guides/discord.md): Configure the Discord channel surface in Loong.
- [Email](https://docs.loongclaw.ai/use-loong/channel-guides/email.md): Configure the Email channel surface in Loong.
- [Feishu/Lark](https://docs.loongclaw.ai/use-loong/channel-guides/feishu.md): Configure the Feishu/Lark channel surface in Loong.
- [Google Chat](https://docs.loongclaw.ai/use-loong/channel-guides/google-chat.md): Configure the Google Chat channel surface in Loong.
- [iMessage](https://docs.loongclaw.ai/use-loong/channel-guides/imessage.md): Configure the iMessage channel surface in Loong.
- [Channel Guides](https://docs.loongclaw.ai/use-loong/channel-guides/index.md): One setup page per actionable Loong channel surface.
- [IRC](https://docs.loongclaw.ai/use-loong/channel-guides/irc.md): Configure the IRC channel surface in Loong.
- [LINE](https://docs.loongclaw.ai/use-loong/channel-guides/line.md): Configure the LINE channel surface in Loong.
- [Matrix](https://docs.loongclaw.ai/use-loong/channel-guides/matrix.md): Configure the Matrix channel surface in Loong.
- [Mattermost](https://docs.loongclaw.ai/use-loong/channel-guides/mattermost.md): Configure the Mattermost channel surface in Loong.
- [Nextcloud Talk](https://docs.loongclaw.ai/use-loong/channel-guides/nextcloud-talk.md): Configure the Nextcloud Talk channel surface in Loong.
- [Nostr](https://docs.loongclaw.ai/use-loong/channel-guides/nostr.md): Configure the Nostr channel surface in Loong.
- [Signal](https://docs.loongclaw.ai/use-loong/channel-guides/signal.md): Configure the Signal channel surface in Loong.
- [Slack](https://docs.loongclaw.ai/use-loong/channel-guides/slack.md): Configure the Slack channel surface in Loong.
- [Synology Chat](https://docs.loongclaw.ai/use-loong/channel-guides/synology-chat.md): Configure the Synology Chat channel surface in Loong.
- [Microsoft Teams](https://docs.loongclaw.ai/use-loong/channel-guides/teams.md): Configure the Microsoft Teams channel surface in Loong.
- [Telegram](https://docs.loongclaw.ai/use-loong/channel-guides/telegram.md): Configure the Telegram channel surface in Loong.
- [Tlon](https://docs.loongclaw.ai/use-loong/channel-guides/tlon.md): Configure the Tlon channel surface in Loong.
- [Twitch](https://docs.loongclaw.ai/use-loong/channel-guides/twitch.md): Configure the Twitch channel surface in Loong.
- [Webhook](https://docs.loongclaw.ai/use-loong/channel-guides/webhook.md): Configure the Webhook channel surface in Loong.
- [WeCom](https://docs.loongclaw.ai/use-loong/channel-guides/wecom.md): Configure the WeCom channel surface in Loong.
- [WhatsApp Cloud API](https://docs.loongclaw.ai/use-loong/channel-guides/whatsapp.md): Configure the WhatsApp Cloud API channel surface in Loong.
- [WhatsApp Personal](https://docs.loongclaw.ai/use-loong/channel-guides/whatsapp-personal.md): Configure the QR-linked personal WhatsApp bridge surface in Loong.
- [Channel Recipes](https://docs.loongclaw.ai/use-loong/channel-recipes.md): Tutorial-style paths for gateway-supervised service channels, standalone native-serve lanes, outbound-only sends, and multi-channel rollout.
- [Channel Setup](https://docs.loongclaw.ai/use-loong/channel-setup.md): Practical setup guidance for gateway-supervised, standalone native-serve, plugin-backed, and outbound-only delivery surfaces.
- [Channels](https://docs.loongclaw.ai/use-loong/channels.md): Understand the current public delivery surfaces and how to think about them.
- [Common Setups](https://docs.loongclaw.ai/use-loong/common-setups.md): End-to-end setup playbooks that combine providers, channels, and gateway ownership.
- [Configuration Patterns](https://docs.loongclaw.ai/use-loong/configuration-patterns.md): Shared public config shapes for providers, channels, account selectors, memory, and outbound trust.
- [Configuration Reference](https://docs.loongclaw.ai/use-loong/configuration-reference.md): Field-level public reference for Loong's TOML config, validation semantics, providers, tools, memory, channels, and Feishu integration.
- [Gateway And Supervision](https://docs.loongclaw.ai/use-loong/gateway-and-supervision.md): Choose the right ownership model for longer-lived gateway-supervised service channels.
- [Gateway Rollout](https://docs.loongclaw.ai/use-loong/gateway-rollout-playbook.md): Move from one healthy service channel to explicit multi-channel gateway ownership.
- [Local Model Plus Outbound Delivery](https://docs.loongclaw.ai/use-loong/local-outbound-playbook.md): Use local or self-hosted inference while keeping delivery truthful as outbound-only.
- [Memory Profiles](https://docs.loongclaw.ai/use-loong/memory-profiles.md): Choose how Loong preserves continuity without turning memory into a second authority.
- [Use Loong](https://docs.loongclaw.ai/use-loong/overview.md): Public overview of the current operator-facing runtime surface.
- [Anthropic](https://docs.loongclaw.ai/use-loong/provider-guides/anthropic.md): Configure the Anthropic provider profile in Loong.
- [Bailian Coding](https://docs.loongclaw.ai/use-loong/provider-guides/bailian-coding.md): Configure the Bailian Coding provider profile in Loong.
- [Bedrock](https://docs.loongclaw.ai/use-loong/provider-guides/bedrock.md): Configure the Bedrock provider profile in Loong.
- [BytePlus](https://docs.loongclaw.ai/use-loong/provider-guides/byteplus.md): Configure the BytePlus provider profile in Loong.
- [BytePlus Coding](https://docs.loongclaw.ai/use-loong/provider-guides/byteplus-coding.md): Configure the BytePlus Coding provider profile in Loong.
- [Cerebras](https://docs.loongclaw.ai/use-loong/provider-guides/cerebras.md): Configure the Cerebras provider profile in Loong.
- [Cloudflare AI Gateway](https://docs.loongclaw.ai/use-loong/provider-guides/cloudflare-ai-gateway.md): Configure the Cloudflare AI Gateway provider profile in Loong.
- [Cohere](https://docs.loongclaw.ai/use-loong/provider-guides/cohere.md): Configure the Cohere provider profile in Loong.
- [Custom](https://docs.loongclaw.ai/use-loong/provider-guides/custom.md): Configure the Custom provider profile in Loong.
- [DeepSeek](https://docs.loongclaw.ai/use-loong/provider-guides/deepseek.md): Configure the DeepSeek provider profile in Loong.
- [Fireworks](https://docs.loongclaw.ai/use-loong/provider-guides/fireworks.md): Configure the Fireworks provider profile in Loong.
- [Gemini](https://docs.loongclaw.ai/use-loong/provider-guides/gemini.md): Configure the Gemini provider profile in Loong.
- [Groq](https://docs.loongclaw.ai/use-loong/provider-guides/groq.md): Configure the Groq provider profile in Loong.
- [Provider Guides](https://docs.loongclaw.ai/use-loong/provider-guides/index.md): One setup page per built-in Loong provider kind.
- [Kimi](https://docs.loongclaw.ai/use-loong/provider-guides/kimi.md): Configure the Kimi provider profile in Loong.
- [Kimi Coding](https://docs.loongclaw.ai/use-loong/provider-guides/kimi-coding.md): Configure the Kimi Coding provider profile in Loong.
- [llama.cpp](https://docs.loongclaw.ai/use-loong/provider-guides/llamacpp.md): Configure the llama.cpp provider profile in Loong.
- [LM Studio](https://docs.loongclaw.ai/use-loong/provider-guides/lm-studio.md): Configure the LM Studio provider profile in Loong.
- [MiniMax](https://docs.loongclaw.ai/use-loong/provider-guides/minimax.md): Configure the MiniMax provider profile in Loong.
- [Mistral](https://docs.loongclaw.ai/use-loong/provider-guides/mistral.md): Configure the Mistral provider profile in Loong.
- [Novita](https://docs.loongclaw.ai/use-loong/provider-guides/novita.md): Configure the Novita provider profile in Loong.
- [NVIDIA](https://docs.loongclaw.ai/use-loong/provider-guides/nvidia.md): Configure the NVIDIA provider profile in Loong.
- [Ollama](https://docs.loongclaw.ai/use-loong/provider-guides/ollama.md): Configure the Ollama provider profile in Loong.
- [OpenAI](https://docs.loongclaw.ai/use-loong/provider-guides/openai.md): Configure the OpenAI provider profile in Loong.
- [OpenRouter](https://docs.loongclaw.ai/use-loong/provider-guides/openrouter.md): Configure the OpenRouter provider profile in Loong.
- [Perplexity](https://docs.loongclaw.ai/use-loong/provider-guides/perplexity.md): Configure the Perplexity provider profile in Loong.
- [Qianfan](https://docs.loongclaw.ai/use-loong/provider-guides/qianfan.md): Configure the Qianfan provider profile in Loong.
- [Qwen](https://docs.loongclaw.ai/use-loong/provider-guides/qwen.md): Configure the Qwen provider profile in Loong.
- [SambaNova](https://docs.loongclaw.ai/use-loong/provider-guides/sambanova.md): Configure the SambaNova provider profile in Loong.
- [SGLang](https://docs.loongclaw.ai/use-loong/provider-guides/sglang.md): Configure the SGLang provider profile in Loong.
- [SiliconFlow](https://docs.loongclaw.ai/use-loong/provider-guides/siliconflow.md): Configure the SiliconFlow provider profile in Loong.
- [Step Plan](https://docs.loongclaw.ai/use-loong/provider-guides/step-plan.md): Configure the Step Plan provider profile in Loong.
- [StepFun](https://docs.loongclaw.ai/use-loong/provider-guides/stepfun.md): Configure the StepFun provider profile in Loong.
- [Together](https://docs.loongclaw.ai/use-loong/provider-guides/together.md): Configure the Together provider profile in Loong.
- [Venice](https://docs.loongclaw.ai/use-loong/provider-guides/venice.md): Configure the Venice provider profile in Loong.
- [Vercel AI Gateway](https://docs.loongclaw.ai/use-loong/provider-guides/vercel-ai-gateway.md): Configure the Vercel AI Gateway provider profile in Loong.
- [vLLM](https://docs.loongclaw.ai/use-loong/provider-guides/vllm.md): Configure the vLLM provider profile in Loong.
- [Volcengine](https://docs.loongclaw.ai/use-loong/provider-guides/volcengine.md): Configure the Volcengine provider profile in Loong.
- [Volcengine Coding](https://docs.loongclaw.ai/use-loong/provider-guides/volcengine-coding.md): Configure the Volcengine Coding provider profile in Loong.
- [xAI](https://docs.loongclaw.ai/use-loong/provider-guides/xai.md): Configure the xAI provider profile in Loong.
- [Z.ai](https://docs.loongclaw.ai/use-loong/provider-guides/zai.md): Configure the Z.ai provider profile in Loong.
- [Zhipu](https://docs.loongclaw.ai/use-loong/provider-guides/zhipu.md): Configure the Zhipu provider profile in Loong.
- [Provider Recipes](https://docs.loongclaw.ai/use-loong/provider-recipes.md): Tutorial-style setup paths for hosted APIs, compatible gateways, local runtimes, and multi-profile configs.
- [Providers And Models](https://docs.loongclaw.ai/use-loong/providers-and-models.md): Choose a provider, understand model selection, and keep provider truth explicit.
- [Skills](https://docs.loongclaw.ai/use-loong/skills.md): Install, expose, and troubleshoot skills without guessing.
- [Tool Surface](https://docs.loongclaw.ai/use-loong/tool-surface.md): How Loong keeps runtime-visible tools truthful and governed.
- [Tools And Memory](https://docs.loongclaw.ai/use-loong/tools-and-memory.md): Public overview of governed capabilities and continuity behavior.
- [Volcengine Plus Feishu Or Lark](https://docs.loongclaw.ai/use-loong/volcengine-feishu-lark-playbook.md): End-to-end rollout from Ark-family hosted provider to a shipped Feishu or Lark runtime.
- [WeCom Rollout](https://docs.loongclaw.ai/use-loong/wecom-rollout-playbook.md): End-to-end rollout for the shipped WeCom AIBot long-connection runtime.

## OpenAPI Specs

- [openapi](https://docs.loongclaw.ai/api-reference/openapi.json)