// For AI Agents
My AI clone accepts your payment.
I built a second brain: an AI fed on everything I've written, recorded, and worked through on coaching, AI, and building businesses. Humans can talk to it at /ask. Agents can now query it directly.
No API key, no account, no signup form. Send a question, pay $0.10 in USDC on Base, get an answer back with citations. The payment flow runs on x402, so any agent with an x402-aware HTTP client can pay and get answered in the same request.
// How it works
- 01
Send a plain POST request with your question. No headers, no auth, no payment yet.
- 02
You get back
HTTP 402with the payment requirements: amount, asset, network, and where to send it. - 03
Your x402 client signs and attaches payment (USDC on Base), then retries the same request with an
X-PAYMENTheader. - 04
You get a 200 with the answer, its citations against my corpus, a confidence score, and a verification flag.
// Product catalogue
Everything below is one request away. Free products respond directly. Paid products reply 402 with the payment requirements, then 200 once your client attaches a valid X-PAYMENT header.
/api/agent/catalog, and a plain-text guide for LLMs is at /llms.txt.Data
Machine-readable indexes agents can use to find what else is here.
Product catalogue
FreeGET /api/agent/catalogFree machine-readable index of every agent-payable product on this site.
Response shape
{
"service": "string",
"network": "base | base-sepolia",
"products": [
{
"slug": "string",
"method": "string",
"resource": "string",
"priceUsd": "string"
}
]
}Inference
Questions answered live by my AI clone, grounded in my own corpus.
Ask the AI clone
$0.10 USDCPOST /api/agent/askOne question to Verbatim, an AI clone answering on behalf of the site owner.
Example request
{
"question": "What is your view on sovereign AI for small business?"
}Response shape
{
"answer": "string",
"citations": [
{
"chunk_id": "string",
"capture_title": "string",
"capture_url": "string"
}
],
"confidence": "number (1-5)",
"verified": "boolean"
}Research brief
$0.50 USDCPOST /api/agent/researchA structured, corpus-grounded research brief in the site owner’s voice: summary, key points, cited evidence, and implications.
Example request
{
"question": "How should an SME sequence its first year of AI adoption?"
}Response shape
{
"brief": "string (markdown, sections: Summary / Key points / Evidence / Implications)",
"citations": [
{
"chunk_id": "string",
"capture_title": "string",
"capture_url": "string"
}
],
"confidence": "number (1-5)",
"verified": "boolean"
}Content
Full-text authored content: notes, articles, and premium frameworks.
Notes index
FreeGET /api/agent/notesFree index of published notes and digests, with the paid resource URL for each.
Response shape
{
"notes": [
{
"slug": "string",
"title": "string",
"type": "string",
"published_at": "string",
"resource": "string"
}
]
}Note (full text)
$0.02 USDCGET /api/agent/note/:slugFull markdown of a single published note or digest.
Response shape
{
"slug": "string",
"title": "string",
"type": "string",
"body": "string (markdown)",
"published_at": "string"
}Insight article (full text)
$0.05 USDCGET /api/agent/article/:slugFull text of a published insight article as structured JSON.
Response shape
{
"slug": "string",
"title": "string",
"description": "string",
"category": "string",
"date": "string",
"url": "string",
"text": "string",
"wordCount": "number"
}Premium framework
$2.00 USDCGET /api/agent/framework/:slugA complete authored framework or playbook document (markdown).
Response shape
{
"slug": "string",
"title": "string",
"version": "string",
"summary": "string",
"body": "string (markdown)",
"updated": "string"
}Tools
Free utility endpoints, no payment required.
Engagement fit recommendation
FreePOST /api/agent/fitFree tool. Submit a business profile (answers to the fit questions) and get a recommended engagement type plus a booking link.
Example request
{
"answers": [
0,
1,
2,
0,
1,
0,
2
]
}Response shape
{
"recommendation": {
"key": "coaching | consulting | howll",
"title": "string",
"description": "string"
},
"booking_url": "string",
"human_page": "string"
}You don't need to hand-roll the payment step. x402 client libraries (for example x402-fetch) wrap your HTTP client, catch the 402, sign the payment, and retry automatically. Point one at any paid endpoint above and it just works.
I built this because owning your intelligence isn't just a line for humans. If agents are going to be the ones asking the questions, they should be able to pay for a real answer the same way a person would, not scrape it, not guess at it. Every answer here comes from my actual corpus, carries citations, and runs through a fabrication check before it goes out. That's the deal whether you're a person on /ask or an agent hitting the API: grounded answers, or an honest hand-off, never an invented one.
