Skip to main content
Glama

Crawl TikTok for a niche we do not cover yet

create_seed_campaign

Starts a keyword-seeded crawl of TikTok to find creators the index does not have yet. Give it hashtags, it walks them, discovers creators, and adds the ones that fit the niche to the index where search_creators can find them.

Use this only after search_creators has come up short: a niche that returns almost nothing at niche_level "wide", or a country where the supply is obviously thinner than it should be. If search already returns hundreds of plausible creators, a crawl will mostly rediscover them and you have spent credits for nothing. Curation beats crawling on any niche that is already covered.

It is not instant. A campaign runs for hours and results arrive gradually, so the shape of the work is: start the campaign, carry on building the list from what is already in the index, then check back with get_seed_campaign later and search again.

Run suggest_campaign_keywords first and seed from what it returns. Guessed tags are reliably worse: asked for bladder health a person writes #pelvicfloor, #womenshealth, #bladderhealth, while the creators themselves post under #kegels, #prolapse, #incontinence and #urinaryincontinence, and #womenshealth is on so many profiles that it steers nothing.

More good seeds is better, not worse. Each one is a frontier row, and the crawl only compounds when a seed turns up high-lift neighbouring tags to follow. Seven guessed tags stall; fifteen mined ones keep going.

THIS SPENDS THE USER'S MONEY, SO IT IS TWO STEPS.

Call it first without confirm. Nothing is started and nothing is charged: you get back exactly what would be crawled, what it costs, and what the user's balance is. Put that in front of the user in your own words and wait for a clear yes. Only then call again with confirm: true and the same arguments.

Never pass confirm: true on the first call, and never infer consent from an earlier "go ahead" about something else. The user agreeing to build a list is not the user agreeing to spend 250 credits on a crawl.

Max 3 campaigns running at once per account.

Tool call template

Replace each placeholder with a value matching the input schema. Only required parameters are included; add optional parameters as needed.

{
  "id": 1,
  "jsonrpc": "2.0",
  "method": "tools/call",
  "params": {
    "arguments": {
      "name": "<name>",
      "keywords": "<keywords>"
    },
    "name": "create_seed_campaign"
  }
}

Input Schema

NameRequiredDescriptionDefault
nameYesWhat this campaign is looking for, e.g. 'UK bladder health creators'.
startNoStart crawling immediately. Default true; false leaves it as a draft.
regionNoISO-3166 alpha-2 country code to bias discovery toward, e.g. 'BG'. A soft preference for ranking the crawl frontier, not a hard filter.
confirmNoSet only after the user has seen the cost and said yes. Omit on the first call to get the confirmation details.
keywordsYesTikTok hashtags without '#', max 20. Use the tags creators actually post under.
max_depthNoHow far to walk from the seed tags. Default 2. Raise for a wider sweep, at proportional cost.
max_profilesNoStop after scraping this many profiles. Default 5000.
min_followersNoSkip creators below this follower count. Default 1000.

Schema Changelog

Changes observed during successful MCP inspections.

  1. First observed

TDQS

A4.8/5.0
Behavior5/5

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

Annotations only mark this as a non-read-only, open-world, non-idempotent, non-destructive operation; the description adds the material behavior an agent cannot infer — that it spends the user's credits, that cost is only revealed on a confirm-less first call, that it is a two-step consent flow with a hard rule against inferring consent, that results arrive over hours rather than instantly, and that three concurrent campaigns is the cap.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Conciseness4/5

Is the description appropriately sized, front-loaded, and free of redundancy?

Well front-loaded: purpose, then when to use, then timing, then seeding advice, then the money/consent protocol. It is on the long side and repeats the consent rule twice ('Never pass confirm: true on the first call' + 'never infer consent'), but that redundancy sits on the highest-stakes behavior, so it is defensible rather than wasteful.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Completeness5/5

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

For an asynchronous, credit-spending, multi-hour campaign with no output schema, the description covers everything an agent needs to act correctly: cost preview flow, waiting behavior, follow-up via get_seed_campaign, seed sourcing, and the concurrency limit. Nothing material is left to inference.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Parameters4/5

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

Schema coverage is 100%, so baseline is 3; the description still adds real semantics on top — that seeds should be mined rather than guessed, that more seeds compound the crawl, and that max_depth scales cost proportionally. It leaves the remaining parameters (min_followers, max_profiles, region, start) entirely to the schema, which already documents them well.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Purpose5/5

Does the description clearly state what the tool does and how it differs from similar tools?

Opening sentence names a specific verb and resource ('starts a keyword-seeded crawl of TikTok') and states the distinguishing purpose: finding creators the index does not have yet. It explicitly contrasts with search_creators, which appears in the sibling list, so the agent can separate the two without opening either schema.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Usage Guidelines5/5

Does the description explain when to use this tool, when not to, or what alternatives exist?

Gives an explicit precondition ('only after search_creators has come up short') with concrete triggers (near-empty niche at niche_level 'wide', thin country supply) and an explicit negative case (search already returns hundreds of plausible creators → crawl wastes credits). It also routes to the correct supporting siblings: suggest_campaign_keywords before, get_seed_campaign after.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

Try in Browser

Glama MCP Gateway

Add one secure layer between your agents and this server.