> ## Documentation Index > Fetch the complete documentation index at: https://tryklint.ai/docs/llms.txt > Use this file to discover all available pages before exploring further. > ## Agent Instructions > Fetch /llms.txt for the complete page index before exploring further. Every docs page is available as markdown by appending .md to its URL. Coding agents should start at /for-agents/coding-agents.md. # Find creators in your niche > Discovers the creators actually publishing in a niche and returns a qualified shortlist. Every name carries its evidence: qualification posts, recent record, and observed brand work. Use when finding influencers or creators in a niche, building a campaign shortlist, or vetting candidates by what they actually publish. This cookbook finds the creators in your niche and comes back with a shortlist your client can interrogate. You don't run it; your agent does. ## Run it Paste this to your agent, filled in: ```text theme={null} Fetch https://tryklint.ai/cookbooks/find-creators.md and execute the cookbook it contains with my context: - niche or brief: [who you need, e.g. professional dog groomers who show their own grooming work on camera] - audience band: [follower range, e.g. 5,000 to 100,000] - location: [optional, e.g. US] - candidates: [optional: handles you already have; the cookbook vets them instead of discovering new ones] ``` 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 Verify you have all of these before calling the API. If any are missing, ask your operator; don't guess. | Input | What it is | Required | Source | | ---------------- | ----------------------------------------------------------------- | ---------------------------- | ---------------------- | | `niche or brief` | The niche to search, or who the campaign needs, as plain language | yes | operator | | `audience band` | Follower range for candidates | no; default 5,000 to 100,000 | operator | | `location` | Declared-location constraint | no | operator | | `candidates` | Existing handles to vet instead of discovering | no | operator | | `KLINT_API_KEY` | API key | yes | operator's environment | ## Objectives Produce exactly one artifact: a shortlist of five to ten creators, each with a dossier in the Output format below, every claim backed by post IDs from your own calls. Nothing else counts as done. ## Procedure Skip this step if the operator supplied `candidates`. Otherwise discover them with `POST https://api.tryklint.ai/v1/instagram/accounts/search`. The account query qualifies on current identity: the profile presented and the content actually published. The posts block demands recent proof, so nobody qualifies on a bio keyword alone. ```json theme={null} { "account": { "query": "[the brief, as identity: e.g. professional dog groomers who demonstrate their own grooming work on camera]", "filters": { "followers_count": { "gte": 5000, "lte": 100000 }, "declared_locations": { "contains_any": [{ "country_code": "US" }] } }, "posts": { "minimum_matching_posts": 3, "filters": { "posted_at": { "gte": "[90 days ago, RFC 3339]" } } } }, "limit": 20 } ``` Drop the `declared_locations` filter if the operator gave no location. Accounts that never declared one can't match it; don't treat that as disqualifying on its own. Per candidate, a handle-scoped sweep. Filters-only searches may sort; `sort` and `query` never appear together. ```json theme={null} { "post": { "filters": { "posted_at": { "gte": "[90 days ago]" } }, "author": { "filters": { "handle": { "in": ["[candidate handle]"] } } } }, "sort": { "by": "posted_at", "order": "desc" }, "limit": 20 } ``` Note cadence, format mix, and the share of posts matching the brief's territory. The same handle scope with `"has_paid_partnership_label": true` added to the post filters. This returns posts Instagram labeled as paid partnerships. It misses deals without the label; say "observed brand work," never "all brand work." Extract the labeled posts to see partners, offers, and how the creator builds sponsored content. For each shortlisted creator, `GET https://api.tryklint.ai/v1/instagram/posts/{id}` on the two or three posts that best prove the brief's requirements. These are the receipts the dossier cites. Check against the Objectives: every creator has qualification evidence with post IDs, a recent-record summary, and a brand-work note. Flags stay flags; you never issue a fraud or authenticity verdict. If fewer than five creators qualify, report the thin result. Offer to widen the band or relax the brief; don't pad. ## Output format Use exactly this structure, one block per creator: ```markdown theme={null} # Shortlist: [brief], [date] ## @[handle], [follower count] - Why they qualify: [one sentence tied to the brief] (posts: [IDs]) - Recent record: [cadence, formats, share of posts on-brief] - Observed brand work: [partners and frequency, or "none observed in window"] - Worth a human look: [anything odd, or "nothing flagged"] ## @[handle] ... ## Scope Searched [brief] at [band], [location]. [Substitutions, thin results, or tracking caveats worth knowing.] ``` ## Boundaries * Qualification rests on the public published record. Audience demographics, story metrics, and rates are not public observations; don't estimate them. * No fraud verdicts. Surface things worth a human look; never stamp an account fake or authentic. * Absence is bounded by what Klint tracks. A creator with no observed brand work may simply predate tracking; the Scope section says so.