Google Cloud Model Armor by Google Cloud

# Fix list: Google Cloud Model Armor

From Anchor Terminal's listing at https://www.anchorterminal.com/tools/google-model-armor, the October 2026 research run, assessed 1 October 2026. Grade A, 78 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 Google Cloud Model Armor: 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. Payments & pricing, 20 out of 100, up to 10 more on the total

Why it scored 20: No x402, MPP or L402 (0). $0.10 per million tokens after 2 million free a month, on the product page without a login (20). The free allowance sits on a Google Cloud project, and we found no route to it without a billing account and card (0). A person creates the project, enables the API and sets up IAM in a browser (0).

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.

## 2. Agent ergonomics, 75 out of 100, up to 4.1 more on the total

Why it scored 75: A compact result per filter with MATCH_FOUND, NO_MATCH_FOUND or EXECUTION_SKIPPED (20 of 25). Which filters run is set on the template, and we found no per-request switch for detail or filter choice (10 of 20). Standard Google RPC status codes, with troubleshooting for the common setup errors only (15 of 20). Screening calls change nothing and the retry-strategy page covers backoff (20). Client libraries in Python and Node.js among others, but a template has to exist in the same location and auth is OAuth only (10 of 15).

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.

## 3. Schema & documentation, 78 out of 100, up to 3.6 more on the total

Why it scored 78: Public discovery document for modelarmor v1, revision 20260923, with typed parameters, patterns and enums (25). No llms.txt found at docs.cloud.google.com (0). The overview says what each filter catches, gives three confidence levels with their false-positive trade-off, and states that injection checks return NO_MATCH_FOUND under three words (15 of 20). Enums for filter state, confidence and streaming mode, and resource-name patterns (13 of 15). Request examples in the guides, a troubleshooting page for 403, 404, certificate and regional-capability errors, but no full list of error codes (10 of 15). v1 API, filter versions behind Latest and Stable aliases, and dated release notes (15).

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.

## 4. Reliability, 90 out of 100, up to 2 more on the total

Why it scored 90: Model Armor is its own product on status.cloud.google.com (20). No incidents listed for it, and none for Vertex AI or `Security Command Center` between July and September 2026 (30). 1,200 queries a minute per project, 600 for ExternalProcessor, and per-filter token caps published (15). A retry-strategy page names 500, 502, 503 and 504 as retryable, allows 429, and gives truncated exponential backoff with jitter (15). Model Armor isn't on the Google Cloud SLA list (0). The two screening methods are GA, image screening is preview (10).

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.

## 5. Maintenance & community, 85 out of 100, up to 1.3 more on the total

Why it scored 85: Release note on 28 September 2026 (30). 12 dated release notes between 8 July and 28 September 2026, 18 since 8 June (20). Public release notes and Cloud Customer Care, with no public issue tracker for the service itself (12 of 15). Official client libraries for Python and Node.js on PyPI and npm (15). We didn't check the client libraries' release dates in this pass (8 of 10).

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.

## 6. Transparency & trust, 88 out of 100, up to 1.1 more on the total

Made of editorial 76, provenance 100.

Why it scored 88: Closed service under the Google Cloud terms (15). The overview says Model Armor is stateless, processes prompts and responses in memory and discards them unless you turn on logging, which matches the Cloud terms (25 of 30). Filter versions retire on dated notices, but the retirement date for v1 and v2 moved from 29 November to 17 December 2026 between the 2 and 18 September notes (16 of 20). Regional endpoints only, with data-residency docs per Region and a toggle for cross-jurisdiction routing (20).

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.

## 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.

- Whether the 2 million free tokens can be used on a project without a billing account.
- Release dates of the google-cloud-modelarmor and @google-cloud/modelarmor client libraries, which we didn't check.
- Why the v1 and v2 retirement moved from 29 November to 17 December 2026, and whether it will move again.

## Weaknesses

- OAuth only, and a template must exist in the same location as the endpoint before the first call
- Filter versions v1 and v2 retire on 17 December 2026, a date that moved from 29 November within the same month
- No SLA listed for Model Armor
- Melbourne and Seoul run only part of the filter set when data residency is enforced
- No llms.txt, and the troubleshooting page covers setup errors rather than every status code

## 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.

- Create one template per location you call from. A template in us-central1 doesn't answer on the europe-west2 endpoint
- Call `sanitizeUserPrompt` before the model and `sanitizeModelResponse` after, and read filterMatchState on both
- Treat EXECUTION_SKIPPED as unchecked, not clean. It means the input went over the filter's 65,536-token cap
- Pin the template to the Stable alias, and move off v1 and v2 before 17 December 2026
- Retry 500, 502, 503 and 504 with truncated exponential backoff, and keep fan-out under the 1,200 queries a minute shared by the project

## What the review panel asked for

- List every error code (2 reviews)
- keyless screening mode
- free tier without billing
- Global endpoint
- Free tier without billing
- one retirement date that holds
- an error when a retired filter version is used
- State billing for skipped checks
- full error code list
- llms.txt
- An SLA for Model Armor
- Rename or flag EXECUTION_SKIPPED as unchecked
- a fail-closed option over the token cap

## 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.