LiveKit Docs MCP Server
Connect Claude Code, Cursor, Codex, VS Code, Gemini CLI, Copilot CLI, and other MCP clients to LiveKit's free documentation MCP server for current realtime voice AI, WebRTC, SDK, CLI, and agent implementation context.
Open the source and read safety notes before installing.
Safety notes
- The LiveKit Docs MCP server is a documentation and public-source-code context server; it is not a connection to your LiveKit Cloud project, rooms, participants, calls, or credentials.
- The endpoint uses Streamable HTTP. A plain browser GET may not behave like an MCP client request, so verify it from an MCP-compatible client.
- One documented tool can submit documentation feedback. Do not let an agent submit feedback that includes secrets, private code, customer data, or unsupported claims.
- LiveKit's APIs and agent SDKs move quickly. Use the MCP server or the matching CLI docs commands immediately before editing production agent, telephony, deployment, or SDK integration code.
Privacy notes
- Docs queries, page paths, code-search terms, and feedback text are sent to LiveKit's hosted documentation service.
- Avoid sending proprietary code, call transcripts, phone numbers, customer identifiers, API keys, LiveKit secrets, SIP credentials, or private incident details in docs-search prompts.
- If an MCP client forwards tool results into a model provider, LiveKit docs excerpts and any user-supplied query text may also be processed by that provider.
- The server can fetch public LiveKit documentation and public LiveKit GitHub repository content; it should not be treated as a private project data source.
Prerequisites
- Claude Code, Cursor, Codex, VS Code, Gemini CLI, Copilot CLI, Google Antigravity, or another MCP client with Streamable HTTP support.
- Network access to https://docs.livekit.io/mcp.
- For LiveKit implementation work, access to the LiveKit project, SDK, or repository you are modifying.
Schema details
- Install type
- cli
- Troubleshooting
- No
- Estimated setup
- 3 minutes
- Difficulty
- beginner
Full copyable content
claude mcp add --transport http livekit-docs https://docs.livekit.io/mcpAbout this resource
Overview
LiveKit Docs MCP Server is LiveKit's official Streamable HTTP MCP endpoint at
https://docs.livekit.io/mcp. It gives coding agents direct access to current
LiveKit documentation, docs-page rendering, SDK code search, changelogs, pricing
context, and public SDK metadata without making the agent scrape the site or
guess from stale training data.
This is most useful when building realtime voice AI agents, WebRTC frontends, telephony flows, LiveKit Agent deployments, or SDK integrations where method names, package versions, and best practices change quickly.
Client Setup
LiveKit documents first-party setup snippets for common coding-agent clients:
| Client | Setup |
|---|---|
| Claude Code | claude mcp add --transport http livekit-docs https://docs.livekit.io/mcp |
| Codex | codex mcp add --url https://docs.livekit.io/mcp livekit-docs |
| VS Code | code --add-mcp '{"name":"livekit-docs","type":"http","url":"https://docs.livekit.io/mcp"}' |
| Cursor | Add {"livekit-docs": {"url": "https://docs.livekit.io/mcp"}} to the MCP configuration |
| Gemini CLI | gemini mcp add --transport http livekit-docs https://docs.livekit.io/mcp |
For any other MCP-compatible client, configure the server URL as
https://docs.livekit.io/mcp with HTTP or Streamable HTTP transport.
Tools
The reference currently documents these tools:
| Tool | Purpose |
|---|---|
get_docs_overview |
Browse the full docs table of contents and page descriptions |
get_pages |
Render one or more documentation pages as Markdown, including supported public LiveKit GitHub URLs |
docs_search |
Search documentation by keyword and return page placement plus snippets |
get_changelog |
Retrieve recent release and changelog entries for LiveKit SDKs and packages |
list_sdks |
List LiveKit SDK repositories with package names |
submit_docs_feedback |
Submit constructive feedback on the documentation |
code_search |
Search public LiveKit GitHub repositories for code identifiers, classes, functions, and examples |
get_pricing_info |
Retrieve LiveKit Cloud pricing context, plan information, and inference pricing assumptions |
The server also exposes resources for LiveKit SDK metadata and the docs
overview. LiveKit's lk docs CLI subcommand is powered by the same docs MCP
surface, so CLI users can use the same workflow from a terminal.
Recommended Agent Workflow
Start with browsing before search:
- Use
get_docs_overviewto understand the current docs structure. - Use
get_pagesfor the specific page before changing implementation code. - Use
docs_searchonly when you do not know the right page. - Use
code_searchfor API names, classes, decorators, examples, and SDK signatures that are not fully covered in docs. - Use
get_changelogwhen behavior changed after an SDK upgrade.
LiveKit's own coding-agent guidance recommends preferring browsing and full pages over snippets because search results only provide excerpts.
Use Cases
- Ask Claude Code to check the latest LiveKit Agents quickstart before editing a voice AI project.
- Let Codex fetch current pages for sessions, tasks, handoffs, MCP toolsets, telephony, deployment, or testing before writing code.
- Search public LiveKit SDK repositories for classes such as
AgentSession, decorators such asfunction_tool, or examples for plugin configuration. - Check release notes when a project has mismatched LiveKit package versions.
- Pair the docs MCP server with LiveKit Agent Skills so the agent has both current source-backed docs and reusable architecture guidance.
Safety and Privacy
Treat this as a public documentation context source. It can help an agent write better LiveKit code, but it does not validate your private LiveKit Cloud state, room configuration, SIP provider setup, phone-number compliance, or production deployment policy.
Do not paste secrets, customer records, call transcripts, private repository code, LiveKit API keys, SIP credentials, or incident details into docs-search queries or feedback submissions. Review generated changes before deploying voice, telephony, recording, billing, or production agent behavior.
Source Review
docs.livekit.io/intro/coding-agents.mddocuments LiveKit coding-agent support for Claude Code, Cursor, Codex, Gemini CLI, docs access, the Docs MCP server, and LiveKit Agent Skills.docs.livekit.io/reference/developer-tools/docs-mcp.mddocuments the server URL, Streamable HTTP transport, client install snippets, tools, and resources.docs.livekit.io/reference/developer-tools/livekit-cli/docs.mddocuments the matchinglk docscommands and confirms they are powered by the LiveKit Docs MCP server.docs.livekit.io/llms.txtpublishes the Markdown-oriented docs index used by coding agents and LLM workflows.docs.livekit.io/agents/logic/tools/mcp.mddocuments how LiveKit Agents can consume MCP servers throughMCPToolset, which is adjacent but distinct from this docs server.
Duplicate Review
Checked current content/mcp/, content/skills/, content/tools/, open pull
requests, and repository-wide content for LiveKit Docs MCP, livekit-docs,
docs.livekit.io/mcp, LiveKit coding agents, and LiveKit MCP server.
Existing LiveKit Agents and LiveKit Agent Skills entries mention the docs MCP
server as companion context, but no dedicated LiveKit Docs MCP entry, exact
source URL duplicate, target file, or open duplicate PR was found.
Disclosure
Editorial listing. No paid placement or affiliate link is used. The LiveKit Docs MCP server is a free documentation endpoint from LiveKit. LiveKit also offers commercial cloud infrastructure for hosted realtime and agent workloads.
Source citations
Add this badge to your README
How it compares
LiveKit Docs MCP Server side by side with 3 alternatives on trust, install, platform support, and disclosed safety notes — all from reviewed registry metadata.
| Field | LiveKit Docs MCP Server Connect Claude Code, Cursor, Codex, VS Code, Gemini CLI, Copilot CLI, and other MCP clients to LiveKit's free documentation MCP server for current realtime voice AI, WebRTC, SDK, CLI, and agent implementation context. Open dossier | Better Icons MCP Server MCP server and Agent Skill from better-auth for searching, recommending, retrieving, batching, and syncing Iconify-powered SVG icons from 200,000+ icons across 150+ collections. Open dossier | jCodeMunch MCP Server Token-efficient MCP server for source-code exploration that indexes codebases with tree-sitter, then lets agents retrieve symbols, file outlines, dependency context, blast radius, dead code, PR risk, and targeted source snippets instead of reading whole files. Open dossier | Langfuse Docs MCP Server for Claude Connect Claude Code, Cursor, Copilot, Windsurf, and other MCP clients to the public Langfuse documentation MCP server for tracing, prompt management, evaluation, and agent observability implementation help. Open dossier |
|---|---|---|---|---|
| Trust | ||||
| Install risk | Review first | Review first | Review first | Review first |
| Notes | Safety ✓ Privacy ✓ | Safety ✓ Privacy ✓ | Safety ✓ Privacy ✓ | Safety ✓ Privacy ✓ |
| Category | mcp | mcp | mcp | mcp |
| Source | source-backed | source-backed | source-backed | source-backed |
| Author | LiveKit | better-auth | J. Gravelle | Langfuse |
| Added | 2026-06-18 | 2026-06-18 | 2026-06-18 | 2026-06-18 |
| Platforms | Claude CodeCodexCursorClaude Desktop | Claude CodeClaude Desktop | Claude CodeClaude Desktop | Claude CodeClaude Desktop |
| Source repo | — | — | — | — |
| Safety notes | ✓The LiveKit Docs MCP server is a documentation and public-source-code context server; it is not a connection to your LiveKit Cloud project, rooms, participants, calls, or credentials. The endpoint uses Streamable HTTP. A plain browser GET may not behave like an MCP client request, so verify it from an MCP-compatible client. One documented tool can submit documentation feedback. Do not let an agent submit feedback that includes secrets, private code, customer data, or unsupported claims. LiveKit's APIs and agent SDKs move quickly. Use the MCP server or the matching CLI docs commands immediately before editing production agent, telephony, deployment, or SDK integration code. | ✓Better Icons can write icons into a project file through `sync_icon`; review the target path, component name, framework, and diff before committing generated code. Icon output can include raw SVG, JSX/component usage, URLs, or generated project code. Review SVG attributes and generated code before use in production UI. Learned preferences and recent icons can influence future recommendations; clear preferences when switching projects or style systems. Batch retrieval can add many icons quickly. Keep requested sets bounded and avoid mixing unrelated icon styles in a shared design system. | ✓jCodeMunch can index source trees and return exact source snippets, symbol metadata, dependency context, git-derived risk signals, and repository structure to the MCP client. Use repository allowlists and trusted-folder controls before giving an agent access to private, customer, regulated, or production codebases. File watching, auto-reindexing, login-service registration, and MCP-client hooks are optional behaviors; review them before enabling persistent background behavior. The upstream license permits non-commercial use for free but requires a paid license for commercial use. Treat generated code summaries and repository analysis as advisory context, not approval to edit, delete, publish, or refactor without review. | ✓The public Docs MCP endpoint is unauthenticated and documentation-only; it should not be treated as a connection to your Langfuse project data. Langfuse also documents an authenticated project MCP server at /api/public/mcp; that separate endpoint exposes read and write tools by default, so restrict write tools if you need read-only access. MCP tool schemas are self-describing and may evolve; clients should inspect live capabilities instead of hard-coding a fixed tool list. Use human review for generated tracing, prompt, evaluation, or production-debugging changes before shipping them. |
| Privacy notes | ✓Docs queries, page paths, code-search terms, and feedback text are sent to LiveKit's hosted documentation service. Avoid sending proprietary code, call transcripts, phone numbers, customer identifiers, API keys, LiveKit secrets, SIP credentials, or private incident details in docs-search prompts. If an MCP client forwards tool results into a model provider, LiveKit docs excerpts and any user-supplied query text may also be processed by that provider. The server can fetch public LiveKit documentation and public LiveKit GitHub repository content; it should not be treated as a private project data source. | ✓Icon search queries, selected icon IDs, preferred collections, recent icons, target icon file paths, framework names, and generated SVG/component code may be visible to the MCP client and model provider. `sync_icon` can expose local project structure and icon-file conventions to the agent. Avoid sending private product names, unreleased feature labels, customer-specific UI terms, or proprietary design-system names in icon search prompts when that context is sensitive. | ✓Local indexes can contain file paths, symbol names, signatures, summaries, byte offsets, source snippets, dependency relationships, and git-derived metadata. Optional hosted summarizers or embedding providers may receive source-derived text when configured; prefer local or disabled summarizers for confidential repositories. The upstream README discloses anonymous aggregate token-savings telemetry with an opt-out setting; review and disable it where policy requires. Agent hooks and config-audit features may inspect MCP client settings and agent instruction files; review outputs before sharing them outside the project. | ✓Queries sent to the public Docs MCP endpoint and search-docs API are processed by Langfuse's hosted documentation service. Avoid including proprietary code, customer data, prompts, traces, or API keys in documentation-search queries. The authenticated Langfuse platform MCP uses Basic Auth with project-scoped public and secret keys; never paste the raw key pair or base64 header into prompts, commits, screenshots, or shared logs. If you connect the platform MCP, observations, prompt content, datasets, scores, comments, metrics, model metadata, and media from that project can be surfaced to the MCP client and model provider. |
| Prerequisites |
|
|
|
|
| Install | | | | |
| Config | | | | |
| Citations | ||||
| Claim | Unclaimed | Unclaimed | Unclaimed | Unclaimed |
Signals
Loading live community signals…
A short, calm digest of reviewed Claude resources. Unsubscribe any time.