Install Jentic One Beta
Jentic One is a self-hosted execution layer for AI agents. It lets your agent call the SavedThat API, or any other public or private API you need. You set the rules, the agent never sees your credentials, and every call is logged.
Two steps, two machines. Install the instance in a safe environment, then register your agent from wherever it runs.
Step 1: Jentic One Host machine
# On the machine that will host your Jentic One instance:
curl -fsSL "https://jentic.com/install.sh?src=apis&api=%2Fapis%2Fsavedthat.app%2Fsavedthat" | shStep 2: Agent machine
# On the machine where your agent runs (keep this separate from the instance):
curl -fsSL "https://jentic.com/install.sh?src=apis&api=%2Fapis%2Fsavedthat.app%2Fsavedthat" | sh
jentic register # connects your agent to your Jentic One instanceJentic One is in public beta. The setup above keeps your agent separate from the instance, which is what you want before using real credentials: an agent running as the same OS user as Jentic One can read its stored keys directly. Just evaluating? A single local install is fine to start. See the secure deployment guide for the tiers.
What an agent can do with SavedThat API.
Save a YouTube, Instagram Reels, or TikTok video by URL for asynchronous transcription and indexing
Search saved video transcripts with a natural-language query using hybrid semantic and full-text ranking
Narrow a transcript search to one platform such as youtube, instagram, or tiktok, and set min_similarity or a limit of up to 50 hits
Jump to the exact moment through each hit's matching chunk text, timestamp, and deep link
GET STARTED
Fetch the full transcript and timestamped chunks of any saved video
Patterns agents use SavedThat API for, with concrete tasks.
★ AI agent recalling something from a saved video
When a user asks their agent about a tip they half-remember from a video, the agent searches Jentic for 'search saved video transcripts', loads the searchTranscripts operation, and sends the question as a natural-language query. It answers with the matching chunk text and a timestamped deep link back to the exact moment, and the user's SavedThat token is injected at call time rather than handed to the agent.
POST /api/search with query 'how often to feed a sourdough starter', platform youtube, and limit 5, then reply with the top hit's chunk text and deep link
Building a searchable video library
Short-form videos are hard to find again once they scroll past. Saving each Reel, TikTok, or YouTube URL queues it for transcription and indexing, and the saved video list shows everything in the library. Once processed, every video becomes searchable by what was said, not just by title.
POST /api/videos with url https://www.tiktok.com/@chef/video/123, then GET /api/videos to confirm it appears in the library
Turning a saved video into notes
For a talk or tutorial worth keeping, the transcript endpoint returns the full text with timestamped chunks. An agent can summarize it, pull out steps, or quote a passage with the time it was said, without re-watching the video.
Call GET /api/videos/{id}/transcript for saved video vid_abc and summarize it into five bullet points with timestamps
4 endpoints — the savedthat api saves social videos from youtube, instagram reels, and tiktok by url, transcribes and indexes them, and lets you search what was said across everything you have saved.
METHOD
PATH
DESCRIPTION
/api/search
Search saved video transcripts
/api/videos
Save a video by URL
/api/videos
List saved videos
/api/videos/{id}/transcript
Get a video's transcript and chunks
/api/search
Search saved video transcripts
/api/videos
Save a video by URL
/api/videos
List saved videos
/api/videos/{id}/transcript
Get a video's transcript and chunks
What agents get from Jentic-routed access to this vendor.
Setup
Wiring SavedThat by hand means obtaining a Supabase-issued JWT through the authorization server the spec advertises, attaching it as a bearer token, and handling asynchronous saves and 429 rate-limit responses. Through Jentic you install once, import SavedThat from the API Directory, store the token once, and your agent calls it.
Permission scoping
SavedThat sends the search query and video URL in the request body, so rules bound which operations your agent may call, such as searching and reading transcripts, not which videos it reaches. You choose the operations it may call, so saving new videos is not included unless you add it.
Credential isolation
Your SavedThat token is stored once, encrypted, by your own Jentic One instance and injected at execution time. It never enters the agent's prompt, logs, or context.
Intent-based discovery
Agents search Jentic by intent such as 'search saved video transcripts' or 'save a TikTok video', and Jentic returns the matching SavedThat operation, such as searchTranscripts or saveVideo, with its input schema.
Alternatives and complements available in the Jentic catalogue.
Specific to using SavedThat API through Jentic.
What authentication does the SavedThat API use?
Per its OpenAPI spec, the SavedThat API uses a Supabase-issued bearer JWT in the Authorization header on all four operations. The spec points to /.well-known/oauth-protected-resource for discovering the authorization server. Through Jentic the token is stored by your own instance and attached at call time, so it stays out of the agent's context.
Which platforms does SavedThat support?
The spec lists YouTube, Instagram Reels, and TikTok URLs for saving, and the search platform filter accepts youtube, instagram, or tiktok.
What are the rate limits for the SavedThat API?
Per the OpenAPI spec, transcript search is limited to 30 requests per minute and saving videos to 10 requests per minute, with 429 returned past those limits. Check https://savedthat.app/llms.txt for any updates.
How soon can I search a video after saving it?
Saving returns 202 Accepted and the video is transcribed and indexed asynchronously. If another user has already transcribed the same video, processing is instant.
Is there a SavedThat MCP server?
You don't need an MCP server to give your agent SavedThat. Jentic connects it directly from the API Directory: import it, store your SavedThat token once, and your agent calls search or save when it needs to.
How do I search my saved videos through Jentic?
Search Jentic for 'search saved video transcripts', load the searchTranscripts operation, and execute POST /api/search with your query, optionally adding platform, limit, or min_similarity. To run it on your own infrastructure, install Jentic One from its GitHub repo.
Can I limit what my agent is allowed to do with the SavedThat API?
Yes. Allow only POST /api/search and GET /api/videos/{id}/transcript, and the agent can look things up in your library and read transcripts without adding new videos to it. Every call it makes is logged.
For Agents
Save YouTube, Instagram Reels, and TikTok videos by URL and search their transcripts by meaning, getting back the matching moment with a timestamp and deep link.
Use for: Save this TikTok so I can search it later, Find the video where someone explained sourdough starter feeding, Search my saved YouTube videos for advice on negotiating salary, Get the transcript of the Reel I saved yesterday
Not supported: Does not download videos, post to social platforms, or transcribe uploaded files. Use for saving social video URLs and searching their transcripts only.
The SavedThat API saves social videos from YouTube, Instagram Reels, and TikTok by URL, transcribes and indexes them, and lets you search what was said across everything you have saved. Search is hybrid semantic and full-text, so a natural-language query finds the moment a topic came up even when the wording differs, and each hit returns the video, the matching chunk text, a timestamp, and a deep link. Full transcripts with timestamped chunks are available for each saved video. Saving is asynchronous and returns immediately when another user has already transcribed the same video.
This API is usable in Jentic One now. Its AI-readiness score against Jentic's framework shows where it stands today and where improvements would make it even easier for agents to use.
Base layer of spec validity and structural soundness.
Aggregated quality score from linter diagnostics, weighted by severity.
Percentage of `$ref` references that resolve successfully.
Checks whether the API description parses successfully and conforms to its declared specification (e.g., OpenAPI).
Structural correctness score based on schema issues using logarithmic dampening.
Clarity, completeness, and ingestion readiness for developers and tooling.
How richly the API is illustrated with examples.
Percentage of examples that conform to their schemas.
Percentage of operations with complete response definitions (success, client error, server error).
Health of API ingestion, bundling, and resolution within Jentic pipelines.
Semantic breadth, depth, and agent comprehension for AI systems.
Coverage of descriptions across API elements.
Coverage of RFC 9457 Problem Details for error responses.
Coverage, uniqueness, and casing consistency of operationIds for AI inference.
Coverage of summaries across operations/tags/info.
Functional utility, complexity comfort, and AI orchestration readiness.
Agent comfort level based on API operational and structural complexity.
Trust, risk posture, and security compliance.
Average quality of security schemes based on authentication method strength (weakest link for OAuth2).
Findability, semantic richness, and reasoning readiness.
Clarity and depth of descriptions across API elements.
Score it yourself
Every API in the directory is allowlisted, so you can re-score it with no key required.
npx @jentic/api-scorecard-cli score <openapi-url>