Dev toolsCAN I VIBECODE
TranscriptAPI?
BUILD FIRST
- One specific customer type and one primary use case
- A vertical slice that completes the main job without manual workarounds
- Visible limits, review steps, and a fallback when an external dependency fails
LEAVE FOR LATER
- Broad third-party integrations
- High-scale collaboration, syncing, and permissions
- Automation that can cause harm without human review
RESEARCH NOTES
What this tool actually does.
TranscriptAPI is a hosted REST API for retrieving YouTube transcripts and YouTube metadata, aimed at developers building search, RAG, monitoring, and content workflows. It exposes endpoints such as /youtube/transcript (with optional language-priority list, 1 credit), /youtube/info (free metadata + available transcript languages), /youtube/search (1 credit), /youtube/channel/resolve (free), /youtube/channel/search (1 credit), /youtube/channel/videos (1 credit/page), /youtube/channel/latest (free, RSS-based), and /youtube/playlist/videos (1 credit/page). Auth uses Bearer API keys. It also ships an MCP server and agent skills so the same features work over REST, MCP, and agent tooling. The service is specifically a YouTube transcript/data retrieval API — its own docs state it does NOT transcribe arbitrary uploaded audio files or live microphone streams, and recommends routing those through
RESEARCHED BUILD PROMPT
Give this to your coding agent.
Build 'TubeTranscriptAPI' — a focused replacement for TranscriptAPI: a hosted REST API plus MCP server that retrieves YouTube transcripts and YouTube metadata. Core workflow: a developer signs up for an API key, then calls GET /youtube/transcript?video_url=... with Bearer auth and optional 'language' priority list to get a timestamped transcript; other endpoints expose /youtube/info, /youtube/search, /youtube/channel/resolve, /youtube/channel/search, /youtube/channel/videos, /youtube/channel/latest, and /youtube/playlist/videos. Implement a credit system: transcripts, search, channel lists, and playlist pages cost 1 credit; info/resolve/latest are free. Technical approach: FastAPI (or Express) with PostgreSQL for API keys and credit balances, Redis for rate limiting, and an internal queue that fetches transcripts via yt-dlp or youtube-transcript-api, caching results keyed by video ID + language in DB/S3. Serve OpenAPI/Swagger docs and an MCP server exposing the same endpoints as tools; include a dev-mode mock that works without YouTube. Explicitly do NOT copy TranscriptAPI's brand, logo, colors, or site design — the UI/docs/README style must be your own. Out of scope: file upload transcription, ASR, live streaming, and any UI beyond a simple API key dashboard and docs. Deliver a README covering setup, env vars, migrations, example curls, credit accounting, and known limitations (YouTube ToS, rate limits, language availability, transcript unavailability).