Lordevo Shop

# Fix list: Lordevo Shop

From Anchor Terminal's listing at https://www.anchorterminal.com/tools/lordevo-shop, the October 2026 research run, assessed 30 October 2026. Grade B, 65.6 out of 100.

This is everything the published grade says the listing lacks, the biggest possible gain to the total first. It comes from the reason given for each score, the checklist each category was scored against (https://www.anchorterminal.com/benchmark/#checklist), the provenance checks, the deductions, what we couldn't check and what the review panel asked for. A fix counts at the next check, once it's public.

For a coding agent working on Lordevo Shop: work through the items below in the product, its docs and its public pages. Each category gives the reason for its score, with the points each checklist item earned, and the checklist itself, so the gap is the items that earned less than their points. Change the product, not the wording, and keep a note of what you changed and where it's published.

## 1. Reliability, 48 out of 100, up to 10.4 more on the total

Why it scored 48: Hosted reading, scored on the HTTP API at shop.lordevo.com. The status page at shop.lordevo.com/status is the vendor's own, rebuilt every 10 minutes from the shop's request log with a JSON twin, but it has no components and the log starts on 29 October 2026 at 12:57 UTC (12 of 20). The record is under a day long. It shows 0 server errors in 15,871 logged requests, 6 process starts and 7 tunnel restarts in 24 hours, one unlogged gap of 12 minutes, and a known issue that an outside registry measured low 30-day uptime on the A2A endpoint, which the vendor traces to tunnel drops. Eight days of operation and no outage of the paid API recorded, but restarts and drops the vendor itself reports (10 of 30). No rate limit numbers for paid calls were found; the only figure is the trial cap of 3 calls per IP per UTC day (3 of 15). 429 and 503 answer with Retry-After and are not charged, the OpenAPI file documents 503 on 63 operations and 429 on the 6 services, every quick check is a read, and a failed call after payment leaves the payment unsettled, so retries are safe. No backoff guidance beyond waiting the given seconds (13 of 15). No SLA. llms.txt says no ranking, traffic or security outcome is promised (0). No beta or preview label; the catalog marks every service on sale and backend ready (10). 12 + 10 + 3 + 13 + 0 + 10 = 48.

The checklist (https://www.anchorterminal.com/benchmark/#checklist-reliability):

Hosted APIs, MCP servers, models and platforms.

- 20, a public status page with component history (Statuspage, Instatus, BetterStack or the vendor's own).
- 0 to 30, the incident record for the last 90 days on that page. 30 for a clean record or trivial incidents only, 20 for minor incidents only, 10 for one major outage (an hour or more of a core API down, or errors across the board), 0 for several. 5 when there's no history we could read, and the note says so.
- 15, rate limits documented with numbers.
- 15, documented 429 or overload handling (Retry-After, backoff guidance), and idempotency keys or safe-retry guidance where writes are involved.
- 10, an SLA published for any paid tier.
- 10, the surface agents use is generally available, not beta or preview.

Local packages, SDKs, frameworks and stdio MCP servers.

- 20, installs from an official package with supported runtimes stated.
- 25, a public CI and test suite, passing on the default branch.
- 0 to 25, open crash or regression issues relative to activity (25 for few and handled, 0 for many, old and unanswered).
- 15, semver discipline and breaking changes called out in a changelog.
- 15, version 1.0 or later, or declared stable.

Protocols are read from their reference implementations, the public facilitators or servers, spec stability and test vectors.

## 2. Security & auth, 45 out of 100, up to 9.6 more on the total

Why it scored 45: No credential exists. Each call carries a signed USDC authorisation for one amount on Base, in a PAYMENT-SIGNATURE or X-PAYMENT header, never in the query string; `trial=1` is a flag, not a secret. The wallet key stays with the agent and nothing can be revoked server side, so the scope is per call but a leaked wallet key is total (20 of 30). Every quick check reads public data, the site fix plan writes nothing to the site, the SEO services need an ownership attestation, and the deep active scan is refused with 403 before any 402 unless the host passed a free DNS TXT or well-known verification. The attestation is self-declared and no tool annotations were read (17 of 20). page-markdown, page-meta, pdf-text, feed-parse, link-preview and the verdict tools return third-party page content, and no prompt-injection guidance was found in llms.txt, the OpenAPI file, the catalog or the client README (2 of 15). Each answer carries `charged`, the trial answer carries `x-lordevo-trial-remaining`, and payments settle on chain where the payer can see them, but with no account there is no per-call log for the operator beyond the aggregate status page (6 of 15). No security.txt on shop.lordevo.com, lordevo.com, seo.lordevo.com or pentest.lordevo.com, no disclosure policy, bug bounty, SOC 2 or ISO 27001, and no advisories (0 of 20). 20 + 17 + 2 + 6 + 0 = 45.

The checklist (https://www.anchorterminal.com/benchmark/#checklist-security):

- 0 to 30, the credential model. 30 for OAuth 2.1 with scopes, or scoped and revocable keys with rotation. 20 for plain revocable API keys. 10 for one all-powerful key. 10 off when a secret can travel in a URL query string as a documented option.
- 0 to 20, read-only or least-privilege modes, and confirmation or approval for destructive actions.
- 0 to 15, prompt-injection posture where the tool returns untrusted content (documented mitigations or guidance). A tool that returns no untrusted content gets 10.
- 0 to 15, audit logs or per-call visibility for the operator.
- 0 to 20, a security programme. security.txt or a disclosure policy, a bug bounty, SOC 2 or ISO 27001, advisories handled in public.

Models are read for retention, whether API data trains models (and whether that's off by default), zero-retention options and certifications. Frameworks for telemetry defaults, approval hooks, guardrails and sandboxing.

## 3. Agent ergonomics, 67 out of 100, up to 5.4 more on the total

Why it scored 67: API reading of the checklist. Responses are single JSON objects with some size controls (`type` on dns-lookup, `limit` 1 to 100 on feed-parse, `max_pages`, `max_tokens`, `count`, `max_matches`) and hard caps per tool such as 60,000 characters on page-markdown, with no field selection. The MCP server would score 5 of 25 on its 50 tools with no toolsets, and its card alone runs to about 32,800 characters of names, titles and descriptions (15 of 25). No pagination, since every call is one shot; per-tool limits of 10 URLs, 40 links, 1,000 sitemap URLs, and a `type` filter on DNS (10 of 20). JSON error bodies with `charged: false`, documented meanings for 400, 402, 403, 422, 429 and 503, and 404 bodies in JSON since 30 October, but no error schema and a known issue where a malformed percent escape shows a Cloudflare 520 (14 of 20). All 48 quick checks are reads, failed calls are never charged, results are cached from 2 minutes to 24 hours by tool, and the npm client caches identical calls locally; no idempotency key on the paid POST services, and readOnlyHint or destructiveHint annotations weren't read (13 of 20). One required parameter per quick check with defaults for the rest, official clients on npm and PyPI plus an elizaOS plugin, and a `maxPriceUsd` cap in the client that defaults to $0.05 (15). 15 + 10 + 14 + 13 + 15 = 67.

The checklist (https://www.anchorterminal.com/benchmark/#checklist-ergonomics):

- 0 to 25, context cost. For MCP, the number and size of the tool definitions (25 for ten or fewer compact tools, 15 for 11 to 30, 5 for more than 30, plus up to 10 back for toolsets, dynamic loading or read-only subsets). For APIs, whether responses can be sized (field selection, limits, summaries).
- 20, pagination, filtering and output-size controls.
- 20, actionable, documented error responses, codes and messages an agent can recover from.
- 20, idempotency or safe retries, and for MCP the `readOnlyHint` and `destructiveHint` annotations.
- 15, sensible defaults, few required parameters, and official SDKs in at least two languages.

Models are read for tool use, structured output, prompt caching, context length, batch and SDKs. Frameworks for how much code and how many defaults a tool-calling agent with MCP needs.

## 4. Transparency & trust, 61 out of 100, up to 3.4 more on the total

Made of editorial 34, provenance 87.

Why it scored 61: Closed service with no terms of its own. shop.lordevo.com/terms and lordevo.com/terms answer 404 and the footer links go to Lordevo SEO's terms of 22 October 2026, which cover scans, Stripe orders and refunds and limit liability to the amount paid, without mentioning the shop, x402 or wallets. No legal entity, registration number or address on any page read. The npm client is MIT (10 of 30). Lordevo SEO's privacy policy of 24 October 2026 is specific where it applies (IP addresses in memory for an hour, no cookies or third-party analytics, page text for content plans sent to Anthropic's Claude, OpenAI's Codex or a local model, encrypted CMS credentials), but it doesn't cover the shop's request log, wallet addresses or settlement through Coinbase CDP, and the shop's statements that its verdict tools use a local model and that report links expire after 30 days sit in llms.txt rather than a policy (12 of 30). No deprecation policy or notice period; the changelog dates two reprices in one day and an unreleased section (4 of 20). Cloudflare, Stripe, Google, Anthropic and OpenAI are named as processors in the privacy policy and Coinbase CDP as the settlement facilitator in llms.txt, with http-ping measured from Malaysia and no data location stated for the shop (8 of 20). 10 + 12 + 4 + 8 = 34.

The checklist (https://www.anchorterminal.com/benchmark/#checklist-transparency):

- 0 to 30, source availability and licence clarity. 30 for open source under an OSI licence, 15 for closed with clear terms, 0 for unclear terms.
- 0 to 30, data handling and retention statements that agree with each other (privacy policy, DPA, retention periods, subprocessors).
- 0 to 20, a deprecation policy or notices with dates.
- 0 to 20, telemetry disclosed with an opt-out (local software), or subprocessors and data locations disclosed (hosted).

The other half of Transparency and trust is the provenance score, computed from checked facts (below). The category score is the mean of the two.

Provenance checks not met in full (half of this category, computed from checked facts):

- Domain age: lordevo.com, registered 2021-05-01 (5 years) (11 of 15)
- security.txt: not found (0 of 10)

## 5. Schema & documentation, 80 out of 100, up to 3.3 more on the total

Why it scored 80: An OpenAPI 3.1.0 file (version 1.9.0) with 62 paths and 71 operations, a 200 response schema and example on every quick check, and `x-price-usd` and `x-payment-info` per operation. It declares no security scheme, which fits a keyless x402 API, and catalog.json adds a JSON Schema input per quick check with `additionalProperties: false`. The MCP server card lists the 50 tools without input schemas and we didn't open a session (23 of 25). llms.txt, a Markdown /start page, catalog.json and examples.json (10). Every description says what the tool does, its limits, cache time and typical latency, and some say what it doesn't do (can-fetch never downloads the body, email-validate does no SMTP probing, ip-info has no geolocation or ASN). The LITE tiers name their full versions. Little guidance on choosing between link-check, http-ping and redirect-chain, and no 'not for' lines (13 of 20). 83 query parameters with 75 required markers and 10 enums, service bodies typed with `const: true` and `minLength`, and only the /mcp and /a2a envelopes free-form. The `trial` parameter and `X-Lordevo-Trial` header appear in llms.txt and /start but in none of the 71 operations (11 of 15). 66 parameter examples and a 200 example per quick check, and 400, 402, 403, 422, 429 and 503 documented with their charging on 63 operations, but error bodies have no schema and the 402 shape is described only in prose (12 of 15). OpenAPI version 1.9.0, a dated changelog from 22 to 30 October 2026 with a JSON twin and an unreleased section, no versioned paths and no deprecation notice period (11 of 15). 23 + 10 + 13 + 11 + 12 + 11 = 80.

The checklist (https://www.anchorterminal.com/benchmark/#checklist-schema):

APIs and MCP servers.

- 25, a machine-readable contract (a public OpenAPI file or similar; for MCP, typed JSON Schema inputs on every tool).
- 10, llms.txt or Markdown docs served for agents.
- 0 to 20, descriptions that say what a tool is for, when to use it and when not to, read from the tool definitions in the source or the API reference.
- 0 to 15, typed inputs with enums, constraints and required fields, and no free-form JSON blobs.
- 0 to 15, examples and documented error responses.
- 15, versioning and a public changelog.

Models are read from the API reference, the OpenAPI file, llms.txt, the structured-output and tool-use docs and the model cards. Frameworks from docs a model can follow, typed interfaces, examples and the API reference.

## 6. Maintenance & community, 76 out of 100, up to 2.1 more on the total

Why it scored 76: The newest changes are the four data utilities, the reliability fixes and the status page, all dated 30 October 2026, with OpenAPI version 1.9.0 (30). Nine dated changelog entries between 22 and 30 October 2026, npm releases 0.1.0 to 0.1.3 between 27 and 29 October, and registry versions 1.0.0, 1.1.0 and 1.1.1 (20). Closed service with a public changelog, an unreleased section and support by email to hello@lordevo.com, which the about page says a human owner answers. No community channel, and the GitHub repository the registry entry names, forbot26/lordevo-x402, answers 404 on the web and the API, as does the user page (7 of 15). In the official MCP registry as com.lordevo/shop under the vendor's domain, with current clients on npm and PyPI (15). The npm package is MIT with Node 20.10 or later, optional peer dependencies on viem and @x402, a `node --test` script and no repository field; no CI is visible because no source repository was found, and the PyPI project page sits behind a bot check so only its release feed was read (4 of 10). 30 + 20 + 7 + 15 + 4 = 76.

The checklist (https://www.anchorterminal.com/benchmark/#checklist-maintenance):

- 0 to 30, time since the last release, or the last published model or API change for a closed service. 30 within 30 days, 20 within 90, 10 within 180, 0 older.
- 20, at least three releases or dated changelog entries in the last 90 days.
- 0 to 25, responsiveness. Issues and pull requests answered on GitHub (the open issues and how recent the replies are). For closed services, a public changelog and a support or community channel that answers, 0 to 15.
- 15, presence in the official MCP registry under a verified namespace (MCP servers), or current official SDKs (APIs and models).
- 10, package health, current dependencies and CI.

Models are read for deprecation notice periods and model churn rather than release counts.

## 7. Payments & pricing, 98 out of 100, up to 0.3 more on the total

Why it scored 98: x402 on every paid endpoint, v2 and v1, USDC on Base (eip155:8453), exact scheme, with a manifest of 64 resources at /.well-known/x402.json and 45 of them in Coinbase's x402 Bazaar under the payTo address. We didn't probe the 402; the 30 October listing check's trial call answered 200 (40). Per-call prices for all 48 quick checks, the digest and the 7 service tiers are published in llms.txt, catalog.json, the OpenAPI file and the manifest, $0.002 to $5 (20). 3 free calls per IP per UTC day on tools priced up to $0.005 with no wallet or card, covering 36 of 48 tools, plus free discovery and sample outputs. Two points withheld because the trial is small and skips the $0.01 and $0.02 tools (18 of 20). No signup of any kind; an agent with a funded Base wallet calls and pays (20). 40 + 20 + 18 + 20 = 98.

The checklist (https://www.anchorterminal.com/benchmark/#checklist-payments):

The published rubric, also on the [x402 page](https://www.anchorterminal.com/x402/).

- 40, a machine payment protocol (x402, MPP or L402) on the tool's own endpoints. 10 to 30 when it covers only some endpoints or only goes through a third party, and the note says which.
- 20, per-call or per-unit pricing published without a login. 10 for public plan-only pricing, 0 for "contact sales" or prices behind a login.
- 20, a free tier or trial that doesn't need a card.
- 20, autonomous onboarding, meaning an agent can get access without a person signing up in a browser (keyless use, x402, a programmatic key API).

Payment platforms and agent wallets rarely charge for their own API over a machine protocol, so the first line has steps for them, and the highest one that applies counts. 40 when x402, MPP or L402 runs on all their own endpoints, 30 when it runs on part of their own API, 25 when their merchants can accept one, 20 for running a facilitator, 15 for paying as a buyer, and 0 when the only protocol is their own. Merchant acceptance sits above a facilitator because the platform's own customers can charge agents through it, while a facilitator settles for sellers who wire up the protocol themselves. The counter-argument (a facilitator does more for the protocol as a whole) has a point. Each note says which step applied.

Open-source software you run yourself is scored on its hosted or paid option if it has one. A free, self-hosted package with nothing to buy gets 20, 20 and 20 for the last three lines, and 0 to 40 for the first only if it ships a payment protocol.

## What we couldn't check

What we couldn't read counted as absent. Publishing it on a page a plain HTTP fetch can read (not only in a browser) lets the next check count it.

- unchecked: the MCP server's live tools/list, its input schemas and any readOnlyHint or destructiveHint annotations. We opened no session; the 50-tool count is from the server card and the 30 October listing check, and the card lists no schemas.
- unchecked: the 402 challenge itself. We made no call; the 30 October listing check's trial GET to dns-lookup answered 200 in under a second with `x-lordevo-trial-remaining: 2`, and the vendor's /start page says the examples were verified on 29 October.
- unchecked: the legal entity. No page on lordevo.com, shop.lordevo.com, seo.lordevo.com or pentest.lordevo.com names a company, registration number or address; the about page says a small, independent online service with a human owner, and the submitter wrote Lordevo Team, Malaysia.
- unchecked: the PyPI project page, which sits behind a bot check. The package's publisher, licence and README on PyPI come only from the release feed (author hello@lordevo.com, four releases from 27 to 29 October 2026).
- unchecked: mcpservers.org, which answers with a bot check, so that listing stays the vendor's claim. PayAPI Market (listed 26 October 2026, 13 endpoints, one route verified) and mcp.tc were read; both are stale against the current 48 tools.
- unchecked: npm and PyPI weekly downloads, GitHub stars. The registry entry names github.com/forbot26/lordevo-x402 as the source, and the repository, the user page and the API all answer 404, so no source, CI, issues or stars could be read.
- The terms and privacy pages that lordevo.com/terms and /privacy were expected to draw by script answer 404; the readable terms and privacy policy are at seo.lordevo.com/terms (22 October 2026) and seo.lordevo.com/privacy (24 October 2026), written for Lordevo SEO. pentest.lordevo.com/terms (25 October 2026) covers Lordevo Security. None covers the shop, x402 or wallet data.
- Settlement through Coinbase CDP is the vendor's statement in llms.txt. The Bazaar lists the shop's resources, which is consistent with the CDP facilitator having seen them, but the facilitator URL isn't in the manifest.
- The status page's 7-day window covers only the 20 hours since the request log began, and its restart lists reach back to 23 and 24 October. Whether the tunnel drops it reports touched paid calls can't be told from it.
- Descriptions on third-party listings and in the registry are stale. The registry's 1.1.1 entry says 33 paid checks from $0.01, the npm description says 44 tools, and the catalog has 48 at $0.002 and up after the 29 and 30 October changes. Recorded, not deducted.
- Category fit. Lordevo returns no keywords, search volumes, backlinks, rankings or traffic, so the seo category's capabilities don't apply; the listing uses web.fetch, web.extract, email.verification and pdf.extract. It sits in seo because its own framing is SEO and website security and no category covers website trust, DNS, email and security-header checks.
- Lordevo Shop doesn't do LocalGhost's or letme's job. Its agent-readiness score grades websites, not MCP servers, and it holds no credentials for anyone. Lordevo SEO's privacy policy names Anthropic's Claude, whose models wrote this grade, as one of the models that writes its content plans; no score depends on that.

## Weaknesses

- Eight days old on 30 October 2026, with a request log that starts on 29 October, 6 process starts and 7 tunnel restarts in 24 hours, and a self-reported uptime problem on the A2A endpoint
- No terms or privacy policy for the shop itself; the footer links to Lordevo SEO's, which don't mention x402, wallets or the shop's request log, and no legal entity is named on any page read
- No rate limit numbers for paid calls, only 429 or 503 with Retry-After, and no SLA
- No security.txt on shop.lordevo.com, lordevo.com, seo.lordevo.com or pentest.lordevo.com, and no disclosure policy or bug bounty
- The MCP server lists 50 tools with no toolsets, and the registry entry names a GitHub repository (forbot26/lordevo-x402) that answers 404
- Doesn't cover the seo category's keyword, backlink, ranking or traffic data; the SEO tools are on-page and technical checks

## What costs an agent a turn today

The notes we give agents before they call it. Each one is a workaround an agent shouldn't need.

- Start with GET /api/quick/site-check?url=... ($0.02) or a $0.002 tool with `trial=1`; the trial allows 3 calls per IP per UTC day on tools priced up to $0.005 and answers 402 once used up.
- Use an x402 v2 client (@x402/fetch with ExactEvmScheme, or x402[httpx,evm] in Python) on Base mainnet; the 402 carries v2 terms in PAYMENT-REQUIRED and v1 terms in the body.
- Treat 429 and 503 as a wait for the Retry-After seconds, never as a billing event; any status of 400 or above after a payment leaves the payment unsettled and the body says `charged: false`.
- For the services under /api/svc, send `url` and `authorization` with `attested: true` and a statement of at least 10 characters; the Deep scan also needs `verification.key` from the free /verify step, and long jobs answer 202 with a result_url to poll every 30 to 60 seconds.
- Over MCP, pass `"trial": true` in the arguments for a free call; a paid call returns the payment requirement as the tool result, then retry with `params._meta["x402/payment"]`.
- Set `maxPriceUsd` in the lordevo-x402 client (default $0.05) and keep result_url and verification_key private; report links expire after 30 days.

## When it's done

Send what changed and where it's published as a dispute (https://www.anchorterminal.com/builders/#disputes, or `POST https://www.anchorterminal.com/api/v1/contact` with `"kind": "dispute"`). Disputes are answered in public, and the listing is checked again by the same checklist. Paying for an audit or a listing claim changes nothing here.