Agentzon

# Fix list: Agentzon

From Anchor Terminal's listing at https://www.anchorterminal.com/tools/agentzon, the October 2026 research run, assessed 10 October 2026. Grade D, 51.4 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 Agentzon: 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, 40 out of 100, up to 12 more on the total

Why it scored 40: No status page at status.agentzon.co, agentzon.co/status or agentzon.co/api/status, each 404 on 10 October 2026, so the status page line scores 0 (0 of 20). No readable incident history exists, which takes the 5-point default (5 of 30). Catalogue reads have published figures, 120 a minute per process and client IP, with Retry-After on 429 (10 of 15). The checkout and order limits have no figure, so the line is not full. 429 handling is documented with Retry-After, and Idempotency-Key covers checkout retries, with retry guidance in agents.md (15 of 15). No SLA and no paid tier (0 of 10). The API is live and takes real payments, with checkoutMode live on all 500 products (10 of 10). Total 40.

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: Read endpoints and checkout creation take no credential (no points for a key model). The order token is a bearer secret that reads one order, never expires and cannot be revoked (15 of 30, between the 10 for one all-powerful key and the 20 for a revocable key). The checkout URL is a payment credential passed in the URL fragment, not the query string, so the 10-point deduction does not apply. Read-only and least-privilege. Catalogue reads and checkout creation cannot take money. The privacy policy says Stripe handles card details and the checkout application never receives them, and the API says never to send them to Agentzon, so an agent cannot pay alone (20 of 20). Prompt injection. The catalogue returns product names and descriptions written by the seller, and no guidance on treating them as data was found (5 of 15). Operator visibility. The buyer has no call log. Agentzon keeps checkout diagnostics for 90 days and aggregate request counters, which the buyer cannot see (5 of 15). Programme. security.txt returns 404, and no bug bounty, certification or advisory page was found (0 of 20). Total 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. Maintenance & community, 5 out of 100, up to 8.3 more on the total

Why it scored 5: No dated API release, changelog or version history was found. The policy pages carry an effective date of 8 October 2026, but those are changes to terms, not to the API, so the time-since-change line scores 0 (0 of 30). No dated changelog entries in the last 90 days (0 of 20). Responsiveness. A support address is published, agentzon-support@merit.systems, but replies were not tested because the brief allows only product listing reads. No changelog (5 of 15 for a closed service). No MCP registry entry under this operator. The official registry lists an unrelated server, xyz.agentzon/agentzon, from agentzon.xyz (0 of 15). No public repository, CI or SDK package was found (0 of 10). Total 5.

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.

## 4. Payments & pricing, 40 out of 100, up to 7.5 more on the total

Why it scored 40: No machine payment protocol (x402, MPP or L402) on any endpoint. UCP is a commerce protocol, and its checkout ends at the same payment page, so it earns nothing here (0 of 40). Item prices are published without a login through GET /api/products. 466 priced items from USD 0.99 to 149.99 on 10 October 2026. Prices exclude sales tax, which is added at checkout (20 of 20). No free tier, trial or test mode was found. Every product reports checkoutMode live (0 of 20). No account, registration or key is needed to search or to create a checkout, so an agent gets access without a signup flow (20 of 20). Total 40. The buyer still enters card details by hand, which the security note covers.

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.

## 5. Transparency & trust, 45 out of 100, up to 4.8 more on the total

Made of editorial 35, provenance 55.

Why it scored 45: Closed service with published, dated terms that are clear on the merchant of record, the refund window and the exclusions (15 of 30). The privacy policy names the seller and Stripe, and says retention is as reasonably needed, with a 90-day period only for checkout diagnostics. Retailers and fulfilment providers are named by type, and the seller has no registered address on agentzon.co (12 of 30). No deprecation policy or dated notices were found (0 of 20). Telemetry is disclosed, including Vercel Web Analytics and checkout diagnostics such as IP address, screen size and automation signals, but no opt-out is stated. Stripe and Vercel are named, with no subprocessor list and no data locations (8 of 20). Total 35.

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: agentzon.co, no registry record we could read (0 of 15)
- Status page: not found (0 of 10)
- Changelog: not found (0 of 10)
- security.txt: not found (0 of 10)

## 6. Agent ergonomics, 78 out of 100, up to 3.6 more on the total

Why it scored 78: Responses can be limited (limit 1 to 100, offset) but not reduced field by field. The first 20 items of the unfiltered catalogue came to 15.5 KB, because each item carries its details (15 of 25). Pagination with offset and nextOffset, filters for category and brand, and four sort orders (20 of 20). Errors carry HTTP codes and a short message, and agents.md gives a recovery step for 400, 415, 409, 429 and network failure, but there are no error codes in the body (15 of 20). Idempotency-Key is required on checkout and catalogue suggestions, with safe-retry guidance (20 of 20). Sensible defaults (limit 20, sort featured) and one required field at checkout. No official SDKs, so the SDK half of the line is unmet (8 of 15). Total 78.

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.

## 7. Schema & documentation, 83 out of 100, up to 2.8 more on the total

Why it scored 83: OpenAPI 3.1.0 at /openapi.json, valid JSON, with request and response schemas for every route (25 of 25). llms.txt at /llms.txt and agents.md in Markdown, both linked from the home page (10 of 10). The descriptions say when not to act. Creating a session is not payment, a paid status confirms payment and not shipment, a null price is unavailable and never free, and purchase needs the buyer's authorisation (18 of 20). No MCP tool definitions exist to read. Inputs are typed, with enums for category and sort, length limits, a pattern on Idempotency-Key and required items at checkout (15 of 15). Examples are given in curl and JSON. Responses are listed for 400, 404, 409 and 429, but the error body is one text field with no machine-readable code (10 of 15). The info block has version 1.0.0 and UCP is dated, but there is no public changelog and the URLs carry no version (5 of 15). Total 83.

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.

## 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 checkout page at the returned URL, including its order summary, tax calculation and whether Stripe Link is offered. The brief allows only product listing reads, so no checkout was created.
- unchecked: Stripe Link. The founder's lead says checkout supports it. No mention appears in the pages or the storefront script read on 10 October 2026.
- unchecked: the UCP REST endpoint at https://agentzon.co/api/ucp. A GET returned 404 and no POST was sent, so catalogue search and checkout over UCP are untested.
- unchecked: the three WebMCP tools in a browser. They are registered in the storefront script when document.modelContext exists, but the script was read, not run.
- unchecked: a registered address, state of incorporation or company number for Merit Systems Inc or Merit Systems, Inc. Search returned no matching registry record.
- unchecked: the source retailers for the 500 items and whether any of them is Amazon. The footer says the site is not affiliated with Amazon.
- unchecked: support response times, refund timings in practice and delivery times. The support address was not emailed.
- unchecked: the domain registration date. RDAP for .co was unreachable.
- The sale label is applied to all 466 priced items, with no list or comparison price in the API. Whether a discount is shown on the product page was not checked.
- The catalogue count differs by source. llms.txt, openapi.json and the home page say 500 essentials, the API total is 500 and 466 are purchasable, and the rendered home page showed 466. The listing uses 466 for purchasable items.
- There is a name collision. A different product named agentzon (agentzon.xyz, a Solana skill marketplace) is in the official MCP registry as xyz.agentzon/agentzon. The slug agentzon may need a qualifier before this listing is merged.

## Weaknesses

- No status page, security.txt or disclosure contact was found. Each probe returned 404
- The terms name Merit Systems Inc as seller, but agentzon.co gives no registered address, state or company number
- Change-of-mind returns exclude food, medicines and personal hygiene items. Personal care is the largest department, with 107 of 500 items
- Source retailers are not named. The policies say only that third parties source and deliver orders
- The UCP endpoint advertised in the profile answered 404 to a GET, and the POST route was not tested

## 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.

- Search with GET /api/products. Use only items with checkoutAvailable true and a non-null unitAmount, because a null price means unavailable, never free
- Send every item and quantity in one POST /api/checkout, with a new Idempotency-Key for each purchase attempt and the same key only to retry that basket
- Show the buyer the product image from /api/products/{id}/image and get their approval before opening the checkout URL
- Tell the buyer that catalogue prices exclude sales tax, that shipping is U.S. only and that change-of-mind returns exclude hygiene and food items
- Keep the checkout URL and the orderToken private. A paid status confirms payment, not shipment

## 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.