Skip to main content
This cookbook turns video posts into clean text your workflow can use. You don’t run it; your agent does.

​
Run it

Paste this to your agent, filled in:
The URL carries everything your agent needs. Everything below this line addresses the agent executing it.

​
Context for the agent

Klint is the social context API: it searches and extracts Instagram posts and accounts, returned with transcripts, on-screen text, and creative structure already analyzed. Base URL https://api.tryklint.ai. Every request needs Authorization: Bearer $KLINT_API_KEY; never print or echo the key. This page is the source of truth for this procedure; work from the fetched copy, not memory. The full documentation index is at https://tryklint.ai/llms.txt, and the whole API contract is at https://tryklint.ai/for-agents/coding-agents.md. Last updated July 18, 2026.

​
Inputs

​
Objectives

Produce exactly one artifact: a text document with one section per requested post, in the Output format below, with every identifier accounted for. Nothing else counts as done.

​
Procedure

1

Extract each post

For each identifier: GET https://api.tryklint.ai/v1/instagram/posts/{shortcode-or-id}. A shortcode from the URL works directly. A 404 with type not-tracked means the post is outside what Klint has observed. Record it as untracked and continue with the rest. One miss never aborts the batch.
2

Assemble the text layers

Per post, in this order: the caption; the transcript of the speech; the on-screen text in its reading order; the visual descriptions. For carousels, keep media items in their published order and label each. A null layer means Klint observed none of that content for the post; write “none” rather than inventing or summarizing around it.
3

Verify before you report

Check against the Objectives. Every identifier the operator gave appears exactly once, either as a text section or in the untracked list. Quote text verbatim from response fields. Do not paraphrase, translate, or fill in from memory of the post.

​
Output format

Use exactly this structure:

​
Boundaries

  • Quote layers verbatim from the response. No paraphrase, no translation, no cleanup beyond formatting.
  • “None” for a layer is an observation, not a failure; silent or text-free posts legitimately have empty layers.
  • Untracked posts stay untracked in the report. Never substitute your own knowledge of a video Klint hasn’t observed.