CASE · 01
A user wants to understand what a criterion (Recognition / Innovation / Impact / Contribution) actually requires before gathering evidence
Explains one or all four UK Global Talent endorsement criteria — C1 Recognition, C2 Innovation, C3 Impact, C4 Contribution — covering what reviewers want, the evidence that works, and the common mistake for each, grounded in the Handbook.
“I can break down exactly what each of the four UK Global Talent endorsement criteria needs — the evidence that works and the mistake that sinks each — and tailor it to your profile.”
{
"jsonrpc": "2.0",
"id": 1,
"method": "tools/call",
"params": {
"name": "criterion_breakdown",
"arguments": {
"criterion": "C3",
"route": "tech"
}
}
}{
"content": [
{
"type": "text",
"text": "# Endorsement criteria — Tech Nation (Digital Technology)\n\nBreakdown of criterion C3 · Impact.\n\n## C3 · Impact\n\n**What reviewers want:** Measurable outcomes — user, revenue, or scale numbers that quantify the difference your work made.\n\n**Evidence that works:**\n- Quantified business impact: revenue growth, user counts, scale metrics tied to your work.\n- Industry standards, tools, or practices you influenced beyond a single product.\n- Adoption figures a third party can verify, not self-reported estimates.\n\n**Common mistake:** Stating responsibilities (\"led the ranking team\") instead of outcomes (\"lifted conversion 18% across 4M users\").\n\nSource: https://ukglobaltalentvisa.org/handbook/the-application"
}
]
}Rate limit. criterion_breakdown adds a light model call on top of the Handbook-grounded scaffold, so it is capped at 25 calls per IP per 24 hours. The deterministic scaffold still returns if the model call fails. When the cap trips, the JSON-RPC error includes data.rate_limited=true and data.window_hours.