Bring YouTube into your AI conversations with ProClipper. Search for videos, discover a creator’s latest uploads, and retrieve YouTube transcripts directly through your AI assistant. Research content, explore videos, and get the information you need without switching between apps.
Scored across 6 tools
Most tools have clearly distinct purposes: account info, transcript fetching, transcription job polling, video-listing job polling, video listing, and transcription submission. The async submit/poll pairs (list_videos/get_video_list and submit_transcription/get_transcription) and the similarly named get_transcript/get_transcription could be confused at first glance, but descriptions explicitly clarify each role.
All names use snake_case with clear verb_noun structure and consistent verb styles: get_, list_, submit_. Minor deviations exist in the paired naming (get_video_list vs list_videos, get_transcript vs get_transcription), but the overall pattern remains predictable.
Six tools is well-scoped for a bulk transcript service. Each tool maps to a necessary step in the workflow: account check, video listing, listing-job polling, transcription submission, transcription polling, and transcript retrieval.
The core lifecycle is covered: discover videos, submit transcription, poll progress, and fetch transcripts, plus account/credit visibility. Minor gaps include no job cancellation, deletion, or history listing, but agents can work around these for the stated bulk transcript purpose.
Read account information and remaining transcript credits. Free.
Inputs: no named parameters.
Fetch transcript content for one ready video or all ready results in a completed job. Free. Use a video id returned by get_transcription, or all once combinedReady is true.
Inputs: jobId (required), format, videoId, timestamps.
Get job progress, ready video IDs, skipped videos, and remaining credits. Free. Poll every few seconds until completed, partial, or failed.
Inputs: jobId (required).
Check a video-listing job and retrieve its result. Free; poll every few seconds while pending or processing. Does not submit a new listing or spend credits.
Inputs: jobId (required).
List video metadata from one YouTube channel or playlist, or inspect one video with mode full. Default flat mode is fast; choose full for an exact upload date when available. Use a known video URL for a specific-video question; if only a title is known, find its URL in a flat channel listing first. Costs 1 transcript credit per successful request, even an empty list. Use a channel URL and a small maxVideos to see its latest Videos-tab uploads; excludes Shorts and live tabs. Does not transcribe. If pending or processing is returned, poll get_video_list with jobId instead of submitting again.
Inputs: url (required), mode, maxVideos.
Start an asynchronous transcription job. Costs 1 transcript credit per video checked, including unavailable captions. Processing failures are refunded. Returns jobId; poll get_transcription instead of submitting again.
Inputs: links (required), language, maxVideos, fallbackToOriginal.
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
Loading tool schema changelog…
Add one secure layer between your agents and this server.