How to connect YouTube to OpenClaw

The ClawHub skill teaches OpenClaw about YouTube, but the calls run through the ClawLink plugin and your connected account. Make sure YouTube is connected in the dashboard, then start a fresh chat so OpenClaw reloads the tool catalog. If OpenClaw runs as a persistent gateway, restart it so the new tools register.

Reconnect YouTube from the dashboard, then start a fresh chat if the runtime still has the old tool catalog loaded.

YouTube tool schemas load on demand the first time a tool runs and are cached after that, so this error usually clears on its own: wait a few seconds and retry the same request. If every YouTube call keeps failing with it in a fresh chat, reconnect from the dashboard, and contact support if it still persists — that pattern points to a configuration problem on our side, not something you can fix by reconnecting again.

Two usual causes. The connected account may not have access to the specific workspace, inbox, store, or project in the request — check that first. If access looks right, the agent may have sent a placeholder value (like "YOUR_ID" or an example id from documentation) instead of a real one: ask it to run a list or search tool first, then retry the action with a real id from those results. Most failures at this stage are one of these two, not ClawLink bugs.

First, the disambiguation, because search results for this phrase plus OpenClaw mostly return LLM rate-limit posts: this is YouTube Data API quota, not your OpenClaw or model-provider rate limit. A 429 from your model provider stops the agent mid-thought; this error comes back inside a tool result while the agent is otherwise fine. Reconnecting does not help and neither does waiting a few minutes — Google's quota resets at midnight Pacific Time. Per Google's current documentation a project gets 100 search.list calls, 100 videos.insert calls, and 10,000 units a day for the remaining endpoints, and searching burns its own counter fastest. The fix is to change the shape of the work rather than retry it: search once, keep the ids, then use youtube_get_video_details_batch for many videos in one call and youtube_list_channel_videos to enumerate a channel without touching the search allocation. Note that most guides and AI answers still quote Google's older table, which charged 100 units per search and 1,600 per upload; the numbers here are the ones on Google's getting-started page today.

Ask the agent to diagnose it:

List exactly which YouTube tools you called in the last few minutes and how many times each. Do not call any more YouTube tools until I reply.

On a self-built integration this means your own OAuth client asked for the wrong scopes, and search results answer it with a Google Cloud checklist: add youtube.force-ssl, delete cached tokens, publish the app out of testing mode. On a hosted connection those are not your levers and the cause is usually one of two others. The first is a partial approval: Google's consent screen lets you untick individual permissions, which leaves the connection active but unable to do that class of action, so reads succeed and writes fail. Reconnect and accept the full set. The second is channel scope — the grant is bound to the channel you chose at sign-in, so acting on a different channel returns a permissions error rather than an empty list. There is also one thing you genuinely cannot work around: YouTube supports service accounts only for content owners who manage multiple channels, such as record labels and movie studios, so an ordinary channel has no unattended path and must go through a user consent flow.

Ask the agent to diagnose it:

Run youtube_get_channel_statistics and tell me the exact channel title and id you are connected to. Then run youtube_list_channel_videos and report how many videos you can see. Do not retry the failed call yet.

If you built the integration yourself, check the publishing status of your OAuth consent screen before anything else. Google issues a refresh token that expires after seven days to any Cloud project whose external consent screen is still in Testing, unless the only scopes requested are name, email, and profile — which YouTube's are not. That is why a DIY YouTube agent reliably works for a week and then stops, and the fix is to publish the app, not to re-run the auth flow. On a ClawLink connection there is no consent screen of yours to publish, so this failure shape does not apply; a connection that stops working here is a revoked grant instead, and reconnecting from the dashboard restores it.

Try reconnecting YouTube and complete the consent flow in the same browser session. Partial OAuth approvals or switching accounts mid-flow can leave the connection incomplete.

Yes. ClawLink is the fastest way to connect OpenClaw to YouTube: link your YouTube account once in the browser and OpenClaw can call the YouTube API through 47 ready-made tools — no custom code or token handling.

Paste the setup prompt from this page into OpenClaw. It installs the ClawLink YouTube skill from ClawHub, then you click Connect in the dashboard to authorize YouTube. OpenClaw calls the tools from the next message — no config files or API keys to manage.

About two minutes. Sign in, click Connect next to YouTube in the dashboard, authenticate, and OpenClaw can use it from the next chat message.

The alternative to ClawLink is usually manual OAuth app setup plus your own token handling, permission troubleshooting, and tool plumbing for OpenClaw. That is fine if you want to build and maintain the integration yourself. Most teams just want YouTube working from chat.

Yes, and that is the main reason to use a hosted connection for this provider. Building it yourself means enabling YouTube Data API v3 in a project you own, configuring an OAuth consent screen, and going through Google's verification before anyone outside your own test list can use it — because YouTube's upload and channel-management scopes are sensitive. ClawLink runs the connection through an app that is already registered and verified, so the only screen you see is Google's consent screen asking whether to grant access to your channel. Uploads then run through youtube_upload_video or youtube_multipart_upload_video like any other tool call.

No, and this is worth separating because the two get conflated constantly. It means there is no YouTube or Google API key: no key to generate, no client secret, and nothing added to openclaw.json for this integration. Whatever you already use to run OpenClaw itself — a hosted model provider or a local one — is unchanged by connecting YouTube. The two credentials are unrelated, and so are their limits: OpenClaw's own rate limits and your YouTube Data API quota are separate systems that fail in different ways.

Google allocates quota per project. Per Google's current documentation, a project that enables the YouTube Data API gets 100 search.list calls, 100 videos.insert calls, and 10,000 units a day for every other endpoint. Because the connection runs through a shared registered app rather than a project you own, you are not administering that allocation and cannot raise it from your side. In practice the ceiling people hit first is the search call count, since it has its own counter — an agent refining an open-ended question will search several times per answer. Ask it to search once and then work from the returned ids with youtube_get_video_details_batch.

Yes, but the channel is chosen during the Google sign-in, not afterwards. If your Google login manages more than one channel, Google shows a channel chooser and the grant attaches to whichever one you pick. The symptom of picking wrong is that the agent reports your videos do not exist while the connection itself looks healthy — it is reading a different channel. Reconnect from the dashboard and select the Brand Account on that screen.

The ClawHub skill teaches OpenClaw about YouTube, but the calls run through the ClawLink plugin and your connected account. Make sure YouTube is connected in the dashboard, then start a fresh chat so OpenClaw reloads the tool catalog. If OpenClaw runs as a persistent gateway, restart it so the new tools register.