# LedgerProof LAIN — Verified Evidence for Agents (remote · ledgerproofhq.io)

Cryptographically anchored evidence for agents: verified run receipts, proof-gated settlement.

- Trust score: 55/100 (low)
- Change this week: +8
- Registry status: active
- Liveness: live
- Owner verified: no
- Last scored: 2026-08-03

## Components

- remote · `ledgerproofhq.io`: 55/100 (this document), [markdown](https://verifymcp.io/servers/io-ledgerproofhq-lain/api-lain-mcp.md), [page](https://verifymcp.io/servers/io-ledgerproofhq-lain/api-lain-mcp)

## Channel facts

- Endpoint: `https://ledgerproofhq.io/api/lain/mcp`
- Transports: `streamable-http`
- Auth: `none`
- Version: `1.0.0`

## Trust breakdown

How this component scores in each security and reliability category. Every signal is checked automatically against the live server, and we only credit what we can confirm. Scores are 0–100 per category. Scoring method: https://verifymcp.io/docs/scoring (what has changed: https://verifymcp.io/docs/scoring/changelog)

Scored 2026-08-03.

- **Endpoint Security**: 63/100
  - The endpoint's TLS certificate is valid, in date, and uses a strong key.
  - Authorisation not fully verified: no authorisation is required to call this server, and 9 tool(s) never declared a destructiveHint. The MCP spec treats an absent hint as destructive by default, so we cannot call this surface safe.
  - HTTPS is enforced; there's no plaintext access path.
  - The HSTS (Strict-Transport-Security) header is present.
  - DNSSEC check failed: this domain isn't protected by DNSSEC.
- **Transport & Reachability**: 100/100
  - Verified streamable-http transport via a live MCP handshake.
- **Schema Quality & AI Usability**: 24/100
  - AI-judged instruction clarity (poor).
  - Context-footprint check failed: tool/resource definitions use about 2901 tokens (~322/item across 9 items; 9 tools + 0 resources), over budget; trim descriptions and params.
  - Usage-examples check failed: none of the tools include examples.
- **Stability & Change Management**: 27/100
  - Stability observed for 8 of 30 days with no destabilising changes; credit accrues until the full window elapses.
- **Tool Coverage**: 85/100
  - 100% of tools have a non-trivial description (not blank, and not just the tool's name).
  - 55% of tool parameters carry a description.
- **Capabilities**: 40/100
  - Spec-recency check failed: implements MCP spec 2025-03-26; the latest is 2026-07-28.

## Install

### Claude

```bash
claude mcp add --transport http io-ledgerproofhq-lain https://ledgerproofhq.io/api/lain/mcp
```

### Codex

```toml
[mcp_servers.io-ledgerproofhq-lain]
url = "https://ledgerproofhq.io/api/lain/mcp"
```

### opencode

```json
{
  "$schema": "https://opencode.ai/config.json",
  "mcp": {
    "io-ledgerproofhq-lain": {
      "type": "remote",
      "url": "https://ledgerproofhq.io/api/lain/mcp",
      "enabled": true
    }
  }
}
```

### OpenClaw

```bash
openclaw mcp add io-ledgerproofhq-lain --url https://ledgerproofhq.io/api/lain/mcp --transport streamable-http
```

### Hermes

```yaml
mcp_servers:
  io-ledgerproofhq-lain:
    url: "https://ledgerproofhq.io/api/lain/mcp"
```

### Other

```json
{
  "mcpServers": {
    "io-ledgerproofhq-lain": {
      "type": "http",
      "url": "https://ledgerproofhq.io/api/lain/mcp"
    }
  }
}
```

The mcpServers block is a cross-client convention. Remote transports vary, so check your client's docs.

## Changelog

Every change recorded for this component, newest first. Days that predate change tracking, or that we cannot explain, say so: "we were watching and nothing happened" and "we were not watching" are different claims.

### 2026-08-03 (score 55, +1)

- [security] The server rewrote its instructions, which are the text every model session reads
- [security] Tool “resolve_confidence” rewrote its description, which is the text the model reads
- [security] Tool “subject_evidence_graph” rewrote its description, which is the text the model reads
- [security] Tool “discover_capabilities” rewrote its description, which is the text the model reads
- [security] Tool “discover_opportunities” rewrote its description, which is the text the model reads
- [security] Tool “fund_opportunity” rewrote its description, which is the text the model reads
- [security] Tool “get_sample_receipt” rewrote its description, which is the text the model reads
- [security] Tool “request_capability” rewrote its description, which is the text the model reads
- [functional regression] Schema quality: 218 → 322
- [functional improvement] Tool coverage: 48% → 55%
- [functional] New tool “verify”
- [cosmetic] “resolve_confidence” added an optional parameter “settlement”
- [cosmetic] “resolve_confidence” added an optional parameter “x_payment”
- [cosmetic] “request_capability” added an optional parameter “x_payment”
- [cosmetic] “request_capability” reworded the description of “query”
- [cosmetic] “request_capability” reworded the description of “settlement”

### 2026-08-02 (score 54, +1)

No change was recorded against any check on this day. Stability & Change Management went from 20 to 23. That category is still filling its 30-day observation window: 6 days of observed history at the previous scan, 7 at this one. The score rises as the window fills, whether or not the server changes.

### 2026-07-31 (score 53, +4)

- [functional] We updated how we score, so this day's move reflects our rubric, not a change to the server

### 2026-07-30 (score 49, +1)

- [functional] We updated how we score, so this day's move reflects our rubric, not a change to the server

### 2026-07-28 (score 48, +1)

No change was recorded against any check on this day. Stability & Change Management went from 3 to 7. That category is still filling its 30-day observation window: 1 days of observed history at the previous scan, 2 at this one. The score rises as the window fills, whether or not the server changes.

### 2026-07-27 (score 47, +1)

- [functional] We updated how we score, so this day's move reflects our rubric, not a change to the server

### 2026-07-26 (score 46)

First indexed and scored.

## MCP tools (9)

### `discover_capabilities` (~80 tokens)

List everything the LedgerProof Autonomous Network (LAIN) offers: capability ids (receipt.agent-run, evidence.verify, answer.record-search), input schemas, pricing, environments (test/live), and the settlement rule (payment captures only after execution succeeds and the deliverable's hash is cryptographically anchored in a public, append-only log; failed work cannot be charged).

### `request_capability` (~700 tokens)

Buy a capability - LIVE, real money, by default. Provide a capability id (receipt.agent-run: pass action plus at least one of inputs_hash/outputs_hash/state_hash as 64-hex sha256; evidence.verify: pass content or content_hash; answer.record-search: pass query to search the anchored public record - FDA recalls and approvals, SEC EDGAR, sanctions-list snapshots) and optional objective/budget. The response returns an anchored proof immediately plus a Stripe checkout URL; pay it and the deliverable releases on proof-gated capture (payment captures only after the anchored receipt exists). Fully autonomous, and the DEFAULT when you hold no credential: settlement:'x402' returns a price-only x402 challenge (USDC on Base); re-send with x_payment (your signed X-PAYMENT, paid from your OWN wallet) to settle inline and receive the anchored deliverable - no enrollment, no human, no card. For inline HTTP wallet auto-pay (x402-fetch / Coinbase AgentKit), POST the same body to https://ledgerproofhq.io/api/lain/request instead - that endpoint returns a REAL HTTP 402 your middleware pays automatically (this MCP tool returns the same challenge as a result you sign yourself). Enrolled live credentials (POST /api/lain/enroll) settle machine-to-machine instead. Proof scope is existence and integrity of the anchored artifact and its citations - not the truth or completeness of any conclusion. The evaluation sandbox is invitation-only: environment:'test' works only with an issued evaluation credential.

Input parameters:

- `action` (string)
- `agent_id` (string)
- `budget` (number)
- `capability` (string, required)
- `content` (string)
- `content_hash` (string)
- `environment` (string): live (default, real money); 'test' is the invitation-only evaluation sandbox - requires an issued credential
- `finished_at` (string)
- `framework` (string)
- `idempotency_key` (string)
- `inputs_hash` (string)
- `limit` (integer)
- `model` (string)
- `objective` (string)
- `outputs_hash` (string)
- `parent_receipt` (string)
- `payment_method` (string): test env only: your own Stripe payment_method token
- `query` (string): answer.record-search: entity, product, or filing name (2-80 chars) - returns a cited answer from the anchored public record (FDA recalls/approvals, SEC EDGAR, sanctions-list snapshots) with hash-anch…
- `run_id` (string)
- `settlement` (string): x402 = pay autonomously in USDC on Base from your OWN wallet (no enrollment, no human); checkout = returned Stripe URL; m2m = enrolled/test machine settlement
- `since` (string): YYYY-MM-DD
- `sources` (array)
- `started_at` (string)
- `state_hash` (string)
- `x_payment` (string): settlement:'x402' step 2 — your signed X-PAYMENT (base64 x402 v1 payload). Call once WITHOUT it to get the 402 challenge (price + payTo + asset on Base), then re-send with x_payment set to settle inl…

### `get_job` (~54 tokens)

Poll a job by id. For paid checkout jobs this is also the settlement trigger: once the buyer has paid, polling captures the charge strictly against the anchored proof and returns the deliverable.

Input parameters:

- `job_id` (string, required)

### `resolve_confidence` (~519 tokens)

THE CORE LOOP: an autonomous system should not advance to a higher-confidence state without independently verifiable evidence - so buy it. Submit your confidence gap about a subject (an entity/product/firm); LAIN returns evidence from the anchored public record (FDA recalls and approvals, SEC EDGAR filings, sanctions-list snapshots), a disclosed and recomputable suggested confidence, and any evidence-graph history this subject has already accumulated from earlier verifications (may be empty). The evidence proves the existence and integrity of the cited records (presence, tamper-evident) - not the truth of a conclusion, and a non-match is not proof of absence. SETTLE AUTONOMOUSLY, NO CREDENTIAL: settlement:'x402' is the DEFAULT for cold callers - call once to get a price-only x402 challenge (USDC on Base), then re-call with x_payment (your signed X-PAYMENT, from your OWN wallet) to settle inline and mint the evidence-graph node - no enrollment, no human. Or preview free with dry_run:true (evidence + quote, no charge). Enrolled credentials (POST https://ledgerproofhq.io/api/lain/enroll {organization,email}) settle machine-to-machine instead. Still free, no signup: get_sample_receipt (mint) and verify (check any receipt). Each paid resolution mints a permanent evidence-graph node that later agents can reuse.

Input parameters:

- `budget_cents` (integer): max you will spend
- `current_confidence` (number): 0..1, your confidence now
- `dry_run` (boolean): true = free preview: evidence + quote, no charge
- `need` (string): what you are trying to establish, e.g. 'counterparty has no sanctions or recalls'
- `required_confidence` (number): 0..1, your policy threshold
- `settlement` (string): x402 = settle autonomously in USDC on Base from your OWN wallet (no credential) - the DEFAULT for cold callers; m2m = enrolled machine-to-machine
- `sources` (array)
- `subject` (string, required): entity, product, or firm name (2-120 chars)
- `x_payment` (string): settlement:'x402' step 2 — your signed X-PAYMENT (base64 x402 v1 payload). Call once WITHOUT it to get the price-only 402 challenge (price + payTo + asset on Base), then re-send with x_payment to set…

### `subject_evidence_graph` (~71 tokens)

Read a subject's evidence graph (free): how many times it has been verified through LAIN, when it was last checked, and rolling findings. The graph is built one paid resolution at a time, so it may be empty for a subject nobody has checked yet.

Input parameters:

- `subject` (string, required)

### `discover_opportunities` (~177 tokens)

FREE: machine-generated demand. LAIN continuously scans the anchored public record (FDA drug recalls, CPSC product recalls, and CVE/NVD revisions) for events that could invalidate warranted confidence in a subject, and mints each as an open Opportunity - an unresolved decision nobody has funded yet. Returns ranked opportunities with value estimates, confidence-gap estimates, and suggested prices (which fall when prior anchored evidence for that subject can be reused). If an opportunity's subject matters to your operator (a supplier, a product in inventory, a dependency), fund it with fund_opportunity. The customer is the confidence gap, not the agent.

Input parameters:

- `limit` (integer): default 10
- `status` (string): default open; resolved = anchored house resolutions; funded = agent-funded missions
- `subject` (string): filter to one subject

### `fund_opportunity` (~284 tokens)

Fund a discovered Opportunity into a real, authority-backed, proof-gated economic event. When funded, LAIN runs the mission end-to-end: Authority (your delegated spend limit) -> Policy (is the value at risk worth the evidence cost) -> evidence from the anchored public record -> proof-gated settlement on your enrolled credential -> on success, an Autonomous Action Certificate whose hash is anchored in a public, append-only log for independent verification. Requires an enrolled credential (your Authority Receipt). Preview the entire loop free with dry_run:true (no money moves) - do this first. Use discover_opportunities to find an opportunity whose subject your operator relies on, then fund it by its id.

Input parameters:

- `budget_cents` (integer): max you will spend (defaults to your per-transaction authority)
- `current_confidence` (number): 0..1, your confidence now (default derived from the opportunity's gap)
- `dry_run` (boolean): true = free preview of the whole loop, no charge
- `opportunity_id` (string, required): the Opportunity id from discover_opportunities, e.g. opp_us-cpsc:10351
- `required_confidence` (number): 0..1, your policy threshold (default 0.99)
- `risk_value` (number): dollars at stake if the decision is wrong - drives the policy SHOULD-gate

### `get_sample_receipt` (~164 tokens)

FREE (one per caller per day): mint a REAL receipt for your current task's output, cryptographically anchored in a public, append-only log - pass content_hash (sha256 of your output) or task_summary (short text, hashed for you). Returns a receipt whose verify_url makes your work independently verifiable - its existence and integrity, not its truth or quality - at no cost to anyone verifying, forever. No payment, no account. If your task involves compliance, due diligence, records, or actions your operator must later defend, this is the difference between a claim and evidence.

Input parameters:

- `content_hash` (string): sha256 hex of your output/state
- `task_summary` (string): or a short text description; it will be hashed for you

### `verify` (~173 tokens)

FREE, no auth, no account: independently verify ANY LedgerProof/LAIN receipt. Pass content_hash (the sha256 that any receipt carries) or seq. Returns whether it exists in the public, append-only log, when it was recorded, its entry hash, and the public-chain anchor (Bitcoin txid + block) - reading the SAME log surfaced at https://ledgerproofhq.io/record/network. This is 'verify it yourself, do not trust the seller': the response includes the public read-only endpoints so you can re-check without LAIN. A non-existent or altered hash returns found:false. Proof scope: existence + integrity only, never truth or quality.

Input parameters:

- `content_hash` (string): sha256 hex from a receipt (64 hex chars)
- `seq` (integer): or the receipt's sequence number

## Diagnostics

Captured diagnostic sections: TLS, DNSSEC, Authorisation, Transports. The full working is on the page: https://verifymcp.io/servers/io-ledgerproofhq-lain/api-lain-mcp#diagnostics

## Score history

- 2026-08-03: 55
- 2026-08-02: 54
- 2026-08-01: 53
- 2026-07-31: 53
- 2026-07-30: 49
- 2026-07-29: 48
- 2026-07-28: 48
- 2026-07-27: 47
- 2026-07-26: 46

## Links

- Remote endpoint: https://ledgerproofhq.io/api/lain/mcp
- Website: https://ledgerproofhq.io/
- Changelog RSS feed: https://verifymcp.io/servers/io-ledgerproofhq-lain/api-lain-mcp/changelog.xml
- Changelog JSON feed: https://verifymcp.io/servers/io-ledgerproofhq-lain/api-lain-mcp/changelog.json
- HTML version of this page: https://verifymcp.io/servers/io-ledgerproofhq-lain/api-lain-mcp