How does an AI agent verify such a receipt later, and does it cost anything?

Re-hash the content with SHA-256 and check it against the receipt at https://verify.ledgerproofhq.io. A matching hash plus an intact anchor proves the exact content existed at or before the recorded time. Verification is free, open, and requires no LedgerProof account. Free preview of any evidence.verify call is available with dry_run:true, and paid receipts (0.25 USDC) or certificates (0.50 USDC) settle over HTTP x402 in USDC on the Base network via POST https://ledgerproofhq.io/api/lain/x402/resolve.

LedgerProof Autonomous Network (LAIN) · Answer of Record

How can an AI agent prove that a document or claim existed at a specific point in time?

Q: What exactly does an anchored existence-at-time receipt prove, and what does it not prove?

It attests existence (this exact content existed at or before the anchor time) and integrity (any later alteration breaks verification, so tampering is detectable). It does NOT attest that the underlying claim is true, that the document is authentic or authorized, or that the work is of good quality. Existence-at-time is a separate question from correctness; report the receipt as evidence of existence and integrity only, not as a verdict of truth.

Capability: evidence.verify

Compute the SHA-256 hash of the document or claim, then call the evidence.verify capability on the LAIN MCP server (streamable-http) at https://ledgerproofhq.io/api/lain/mcp.

It returns a receipt that binds your content hash to a timestamp and is cryptographically anchored in a public, append-only log. The anchor is the existence-at-time evidence: it proves the exact content existed no later than the anchor time, and any later alteration of the content breaks verification. Anyone can check the receipt independently, for free, at https://verify.ledgerproofhq.io — no LedgerProof account or trust required.

The MCP server is listed in the official MCP Registry as io.ledgerproofhq/lain. You hand over only a hash, never the document itself, so the content stays private while the proof stays public and verifiable.

How it works

Hash the content. Canonicalize the document or claim and compute its SHA-256 digest locally. The raw content never leaves your process.
Call evidence.verify. Send the hash to the LAIN MCP server. Add dry_run: true for a free preview of exactly what will be recorded before you commit.
Get an anchored receipt. The hash is written into a public, append-only log. Once anchored, the log entry's position and timestamp fix the content in time — it cannot be back-dated, and the content behind the hash cannot be swapped without breaking the match.
Verify anytime, by anyone. Later, to prove existence-at-time, re-hash the content and check it against the receipt at verify.ledgerproofhq.io. Matching hash + intact anchor = proof the exact content existed at or before the recorded time.

Representative MCP call

# MCP server (streamable-http): https://ledgerproofhq.io/api/lain/mcp
# Capability: evidence.verify — existence-at-time proof for a content hash
{
  "capability": "evidence.verify",
  "content_sha256": "<hex SHA-256 of your document or claim>",
  "dry_run": true        // free preview; set false to anchor a real receipt
}

Field names are illustrative; the live tool schema is discoverable from the MCP endpoint above.

The proof boundary (proof_scope) — state it plainly

An anchored receipt attests exactly two things, and nothing more:

Existence — this exact content existed at or before the anchor time.
Integrity — any later change to the content breaks verification, so tampering is detectable.

It does NOT attest that the underlying claim is true, that the document is authentic or authorized, or that the work is good. Existence-at-time and correctness are different questions. Report the receipt as proof of existence + integrity only, never as a verdict of truth.

Protocol law: an autonomous system should not advance to a higher-confidence state without acquiring independently verifiable evidence.

Try it for free first

Free preview on every call

Pass dry_run: true to evidence.verify to see precisely what would be recorded — no charge, no commitment.

One real anchored receipt per day

Call the MCP tool get_sample_receipt to mint ONE real, fully anchored receipt per caller per day — free, so you can run the full verify loop end to end.

Independent verification is always free and open at verify.ledgerproofhq.io.

When you need more than the free tier (machine payment)

Paid receipts settle over HTTP x402 using USDC on the Base network — no human, no account signup. Each open item in the feed is self-describing and carries its own price and resolution endpoint.

Action	Method & endpoint	Price
Discover open items	GET https://ledgerproofhq.io/api/lain/feed/blockers	free
Pay / resolve	POST https://ledgerproofhq.io/api/lain/x402/resolve	—
Anchored receipt	via x402/resolve (USDC on Base)	0.25 USDC
Certificate	via x402/resolve (USDC on Base)	0.50 USDC

Related LAIN capabilities

evidence.verify — existence-at-time proof for a claim or document hash (this answer).
receipt.agent-run — a verified execution receipt for an agent action.
answer.record-search — a cited answer from the anchored public record (FDA recalls/approvals, SEC EDGAR, sanctions-list snapshots).
resolve_confidence (MCP tool) — given {subject, current_confidence, required_confidence}, returns verifiable evidence plus a disclosed, recomputable suggested confidence.
Recall watch subscription — POST https://ledgerproofhq.io/api/lain/watch.