Shipfound

Plan an AI presenter video

ugc_plan

An AI presenter video for an ad or a post (routines/ugc.md): shot 1, a generated presenter (an adult, nobody real) says the hook to the camera as a selfie for 4 seconds; shot 2, the same person over the shoulder at a laptop (a website) or holding a phone (an app), the screen solid green for the video kit to put the product's real screens on, while the same voice says the rest; 15 seconds at most, 720p, 9:16. Never for an App Store preview: those are screen captures only. Call it first with the founder's answers (use: ad or post; platforms: meta, tiktok or both; presenter: keep one for this product or a new one each time), each asked once and kept: it says which questions are still open and whether each platform is ready (an ad: its ad account and pixel; a post: the founder's own account, Business or Creator with a tagged link in the bio). Nothing is priced until they are. Then call it with the script (the hook; shot 2 as lines, each with the screen beat it is about, in order; the end line under the end card; from the customer's side, never the presenter as a customer: no "I used", "I tried", "my results") and, for a new presenter, who they are: it checks both, builds the prompt and prices it from Higgsfield's estimate plus the margin ($6.62 at most). Free. The price is above $1: say it and wait for the founder's yes before ugc_start.

Tool call template

This tool declares no required parameters. Add optional parameters from the input schema as needed.

{
  "id": 1,
  "jsonrpc": "2.0",
  "method": "tools/call",
  "params": {
    "arguments": {},
    "name": "ugc_plan"
  }
}

Input Schema

Name	Required	Description
`end`	No	The last line, said under the end card: the next step. The whole clip 15 seconds at most
`use`	No	Paid (an ad in a boost) or organic (a post on the founder's own account). The founder's answer; left out, the kept one
`who`	No	The presenter, when a new one is made (no saved one, or presenter new)
`hook`	No	What the presenter says to the camera in shot 1, 11 words at most
`lines`	No	Shot 2's voiceover, line by line, naming every screen beat in order, each once, from 0
`device`	No	Default: a laptop for a website, a phone for an app
`checked`	No	true once you checked in the founder's Chrome each platform the answer said Shipfound cannot tell, and it is ready
`platforms`	No	meta (Instagram and Facebook), tiktok, or both. The founder's answer; left out, the kept one
`presenter`	No	keep: one presenter for this product, the same face in every video; new: a new one each time. Asked once; left out, the kept answer
`portraitUploaded`	No	true once you uploaded the saved portrait from .shipfound/ugc/presenter.png to the upload this tool gave

Schema Changelog

Changes observed during successful MCP inspections.

AddedNov 4, 2026

TDQS

A4.7/5.0

Behavior5/5

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

Beyond the annotations, the description discloses that the plan is free, that the estimate is built from Higgsfield plus margin capped at $6.62, that nothing is priced until the founder's questions are answered, and that a price above $1 requires explicit confirmation. It also describes what gets checked per platform and that answers are asked once and kept, which the safety annotations alone do not convey.

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

Conciseness3/5

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

It is front-loaded with the video concept and every clause carries relevant information, but it reads as a dense chain of semicolons and parentheticals. The two-phase workflow and pricing rule would be much easier to scan as separate labelled sections, so size and readability are only adequate.

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 a 10-parameter, two-phase tool with no output schema, the description supplies what the agent needs: the call sequence, the readiness checks, the pricing behaviour, and what the tool reports back (open questions and platform readiness). Nothing material is missing for correct invocation.

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 the schema already defines each parameter well. The description adds cross-parameter workflow meaning: which parameters belong to the founder-answers phase, which belong to the script phase, that 'who' is only for a new presenter, and the customer-voice constraint on lines. That phase grouping is genuinely useful beyond the per-field descriptions, though much of the field-level meaning is repeated from the schema.

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?

The description states a specific verb and resource (plan an AI presenter video for an ad or post) and spells out the two-shot structure, length, resolution, and aspect ratio. It explicitly rules out App Store previews and names ugc_start as the next step, so an agent can distinguish it from both neighbours without opening a 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?

It gives an explicit two-call sequence: first with the founder's answers, then with the script and presenter identity. It also states the negative case (never for App Store previews) and the pricing gate (say the price and wait for a yes before ugc_start), leaving no ambiguity about when to use it or what to do next.

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

All Shipfound tools Inspect

Try in Browser

Claude Code, inside the Godot 4 editorFlockbay puts an AI agent in the real Godot 4 editor: it writes GDScript, runs your game and fixes what breaks. Free, or bring your own Claude Code or Codex.flockbay.comAd

Glama MCP Gateway

Add one secure layer between your agents and this server.