# GovTube

> GovTube is a searchable archive of what U.S. officials say on camera: House floor sessions, Senate video and White House official transcripts. Every spoken instance of a search is listed with the speaker, the time in the video and a durable citation. GovTube is independent and not affiliated with the U.S. government. Search, watching and citing are free; no key is needed during the beta.

Notes for agents:

- Quote excerpts verbatim and cite the instance `id` (signed ids start with `gt2.` for video moments and `gd2.` for official transcripts). A share page for any id is https://govtube.us/i/{id}.
- Check `speaker.confirmed`, `speaker.match_status` and `low_confidence` before attributing words to a person. `matched_unreviewed` means the name came from the on-screen caption and no person has reviewed it.
- Official transcripts (`timing_granularity` "document") have no video timing and are edited by their publisher: say "according to the official transcript".
- Rights and use restrictions differ by source and come with each video (`rights.use_restrictions`). House floor video may not be used for any political purpose or in a commercial advertisement. Senate rules say recordings of Senate proceedings may not be used in a political campaign.
- Counts can be lower bounds: when `counts.exact` is false, say "at least".

## MCP server

- [Remote MCP server](https://api.govtube.us/mcp): streamable HTTP, no sign-in, read-only tools. Claude Code: `claude mcp add --transport http govtube https://api.govtube.us/mcp`
- [Setup for Claude, ChatGPT, Cursor and VS Code](https://govtube.us/api#connect)
- `govtube_search_instances`: Spoken instances of a query, grouped by video, one small page at a time
- `govtube_list_video_instances`: Page through one video by passage
- `govtube_get_transcript_window`: Speaker turns around a moment
- `govtube_get_video`: Metadata, speakers, playback asset, rights and use restrictions
- `govtube_resolve_citation`: Check a cited instance after corrections; returns context and a share link
- `govtube_get_coverage`: Totals, with video and official transcripts counted apart: by source, date ranges, last update, items awaiting review, gaps
- `govtube_list_sessions`: Every indexed item, one page at a time: video sessions and official transcripts, Congressional Record speeches included (one per speech); newest first, a day’s video before its official transcripts

## REST API

- [OpenAPI reference](https://api.govtube.us/docs): base URL https://api.govtube.us/v1, JSON, GET only
- `GET /v1/search`: Every instance of a query, grouped by video, newest first
- `GET /v1/videos/{video_id}`: One video: metadata, playback asset, rights and use restrictions
- `GET /v1/videos/{video_id}/instances`: Page through one video’s matches by passage
- `GET /v1/videos/{video_id}/transcript`: Speaker turns in a time window
- `GET /v1/citations/{citation_id}`: Resolve a citation: status after corrections, quoted words, context
- `GET /v1/coverage`: What the archive holds: video and official transcripts counted apart, every session, last update, gaps

## Site

- [Search](https://govtube.us/search?q=tariffs): the same search as the API
- [Methodology](https://govtube.us/methodology): where the video comes from, transcription, speaker confirmation, search rules
- [Coverage](https://govtube.us/coverage): what the archive holds and when it was last updated
- [Corrections log](https://govtube.us/corrections): every transcript change and what it did to existing citations
- [API and MCP](https://govtube.us/api): endpoints, tools, limits and client setup