Happy Scribe

@happyscribe list all my completed transcriptions from the last 7 days and show me which ones have subtitles generated

Found 4 completed transcriptions from the past week: 1. "Team standup 2024-01-15" (ID: tr_abc123) — 12min audio, completed Jan 15. Subtitles: Yes (SRT format). 2. "Client demo recording" (ID: tr_def456) — 45min video, completed Jan 14. Subtitles: Not generated. 3. "Podcast episode 47" (ID: tr_ghi789) — 38min audio, completed Jan 13. Subtitles: Yes (VTT format). 4. "Interview - Sarah Chen" (ID: tr_jkl012) — 22min audio, completed Jan 12. Subtitles: Not generated. This is an illustrative summary based on typical Happy Scribe API responses; actual job metadata and formats depend on your account's processing history.

This example shows Happy Scribe's read capabilities for listing and filtering transcription jobs. Requires a valid API key with read permissions. The MCP can surface job status, duration, and subtitle availability, but cannot retrieve the actual transcript text in the list view—you'd need a separate fetch call per job.

@happyscribe create a transcription for this video file at https://example.com/meeting-recording.mp4 and generate subtitles in SRT format

Transcription job created successfully: • Job ID: tr_xyz789 • Media URL: https://example.com/meeting-recording.mp4 • Status: Processing • Subtitle generation: Enabled (SRT format) • Estimated completion: 8-12 minutes The file is now queued for transcription. Happy Scribe will process the audio, generate a timestamped transcript, and create downloadable SRT subtitles. You'll receive a webhook notification when complete (if configured), or you can poll the job status using the ID above. This example assumes the media URL is publicly accessible.

Demonstrates Happy Scribe's core write action: creating a transcription job with subtitle output. The media file MUST be at a publicly accessible URL or a signed URL—Happy Scribe cannot access local files or private cloud storage without proper permissions. Processing time scales with media length; long videos may take 30+ minutes.

@happyscribe take transcription tr_abc123 and create a translation order for Spanish and French, then summarize what languages are now available

Translation order created (Order ID: ord_456xyz) for transcription tr_abc123: • Source language: English (auto-detected) • Target languages: Spanish (es), French (fr) • Order status: Pending confirmation • Estimated delivery: 24-48 hours (human translation) Once you confirm this order, Happy Scribe's translation team will deliver professionally translated transcripts. After completion, your transcription will be available in: - English (original) - Spanish (es) - French (fr) Note: This order is not yet submitted—call the confirm_order tool with ord_456xyz to finalize and begin processing. This illustrative example shows the two-step order workflow.

Shows Happy Scribe's translation workflow paired with AI reasoning to summarize language availability. Translation orders require explicit confirmation (a safety step to prevent accidental billing). Human translation incurs per-minute charges and takes 1-2 days; machine translation is faster but requires different API parameters. The MCP cannot auto-confirm orders—you must explicitly approve.

When Happy Scribe beats manual note-taking for research teams

A 3-person UX research team runs 8-12 customer interviews per sprint and needs searchable transcripts within hours, not days. Happy Scribe wins here because the Create Transcription tool handles the upload-to-text pipeline in one step, and the 24-tool scope means you can script the entire workflow—upload recording, confirm order, retrieve transcript, push to Notion—without leaving Switchy. The API key auth keeps credentials out of individual laptops. The threshold: if your interviews are under 10 minutes or heavily accented, transcription accuracy drops and you'll spend more time correcting than you saved. For teams doing structured 30-60 minute calls in clear audio, this MCP turns a 2-hour manual task into a 10-minute review.

When this MCP makes sense for small content teams

A 2-person marketing team publishes 4-6 product demo videos per month and needs subtitles for accessibility and SEO. Happy Scribe's Create Subtitle tool generates time-coded captions directly from video URLs, which means you skip the export-import dance with Premiere or Final Cut. The catch: you need publicly accessible video URLs before the MCP can process them, so if your videos live in a private S3 bucket without signed URLs, you'll hit a wall. If your CMS or video host already serves public links, this workflow is a 15-minute setup that runs on autopilot. For teams publishing fewer than 2 videos per month, the API key overhead isn't worth it—just use Happy Scribe's web UI.

When translation orders scale better than manual localization

A 5-person support team maintains help docs in 4 languages and updates them twice per quarter. Happy Scribe's Create Translation Order tool takes an English transcript and queues translations into multiple target languages in one API call, which beats coordinating freelance translators or waiting on Google Translate exports. The 24-tool scope includes webhook management, so you can trigger Slack notifications when translations complete. The boundary: if your docs are highly technical or use domain-specific jargon, machine translation will need heavy post-editing, and you lose the time savings. For general support content with standard vocabulary, this MCP cuts a 3-day localization cycle to under 24 hours.