Amazon Bedrock Guardrails by Amazon Web Services

# Fix list: Amazon Bedrock Guardrails

From Anchor Terminal's listing at https://www.anchorterminal.com/tools/amazon-bedrock-guardrails, the October 2026 research run, assessed 1 October 2026. Grade BB, 75.1 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 Amazon Bedrock Guardrails: 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). Per-policy prices per 1,000 text units published without a login (20). No free tier for Guardrails on the pricing page, and an AWS account needs a card (0). A person signs up in a browser and sets up IAM (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. Maintenance & community, 45 out of 100, up to 4.8 more on the total

Why it scored 45: The newest Guardrails changes we found are InvokeGuardrailChecks on 16 June 2026 and Automated Reasoning refinement workflows on 23 June 2026, about 100 days ago (10). Nothing Guardrails-specific in What's New or the document history since 3 July (0). Public document history and What's New, and support through re:Post and AWS Support, but the history lags the announcements (10 of 15). Current SDKs, boto3 1.43.105 on 29 September 2026 (15). SDKs ship near-daily and support Python 3.10 to 3.14 (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.

## 3. Reliability, 80 out of 100, up to 4 more on the total

Why it scored 80: AWS Health Dashboard with per-service, per-region history and RSS feeds (20). The dashboard renders in JavaScript, and the Bedrock feeds for us-east-1 and us-west-2 carried no items when we read them. StatusGator's mirror of the dashboard lists three Bedrock warnings for increased error rates between 24 August and 10 September 2026, none naming Guardrails, so we count minor incidents only (20). Quota numbers are public only in part, 50 ApplyGuardrail calls a second and 200 text units a second for content, PII and word filters in us-east-1 and us-west-2 per a February 2025 announcement, with the rest in the Service Quotas console (10 of 15). The InvokeGuardrailChecks guide says to retry 429 and 503 with exponential backoff and 500 as is (15). The Bedrock SLA promises 99.9 per cent a region but covers "the Amazon Bedrock APIs for models" and doesn't name Guardrails (5 of 10, our call). ApplyGuardrail and InvokeGuardrailChecks carry no preview label (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.

## 4. Transparency & trust, 70 out of 100, up to 2.6 more on the total

Made of editorial 45, provenance 95.

Why it scored 70: Closed service under the AWS Service Terms (15). Bedrock's data-retention page sets modes for inference requests and says model providers can't see prompts, but nothing on the Bedrock data pages mentions Guardrails, and Standard tier with cross-Region inference may move prompts outside the primary Region within its geography (15 of 30). No deprecation policy or dated notices for Guardrails found (0). Regions are listed per tier and the cross-Region page says which geography data stays in (15 of 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.

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

- security.txt: published but past its Expires date (5 of 10)

## 5. Schema & documentation, 92 out of 100, up to 1.3 more on the total

Why it scored 92: The SDKs are generated from AWS's published service models, and the API reference gives every field with type, pattern and enum (25). The user guide has an llms.txt with about 60 guardrail entries and serves .md pages (10). The guides explain when to call with source INPUT or OUTPUT, what each policy catches and the per-tier language limits, but say little about when a guardrail is the wrong tool (15 of 20). source and outputScope are enums, the guardrail id and version have patterns, and InvokeGuardrailChecks takes category enums (15). Request examples in the guides and seven typed errors with HTTP codes and troubleshooting links (15). Guardrails are versioned resources with a DRAFT and numbered versions, and there's a dated document history, though its last Guardrails entry is 19 November 2025 while What's New posted Guardrails launches in April and June 2026 (12 of 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.

## 6. Agent ergonomics, 93 out of 100, up to 1.1 more on the total

Why it scored 93: outputScope INTERVENTIONS keeps the response to what fired, FULL returns every assessment, and usage says how many text units each policy billed (25). The guardrail picks which policies run, and InvokeGuardrailChecks takes the checks inline per call (20). Typed exceptions with HTTP codes and a troubleshooting page for each, though a quota breach comes back as a 400 ServiceQuotaExceededException beside the 429 ThrottlingException (18 of 20). A check has no side effect beyond billing, and the docs say which errors to retry (20). Official SDKs in Python, JavaScript and the other AWS languages, but ApplyGuardrail needs a guardrail built in advance and every call needs SigV4 signing (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.

## 7. Security & auth, 94 out of 100, up to 1.1 more on the total

Why it scored 94: IAM with SigV4, roles and short-lived credentials, and policies can name a single guardrail ARN (30). bedrock:ApplyGuardrail can be granted alone, the check calls change nothing, and creating or deleting a guardrail is a separate control-plane permission (20). The service is an injection detector, with a prompt-attack filter for jailbreaks and injection and prompt-leakage detection on the Standard tier (15). ApplyGuardrail calls are CloudTrail data events on the AWS::Bedrock::Guardrail resource type, but the CloudTrail page doesn't mention InvokeGuardrailChecks (13 of 15). Disclosure policy and a vulnerability disclosure programme on HackerOne, though the aws.amazon.com security.txt expired on 24 September 2026 (4 of 5), no paid bounty found (2 of 5), Bedrock is in AWS's SOC scope (5), public security bulletins (5), so 16 of 20.

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.

## 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.

- The Guardrails quotas for Regions other than us-east-1 and us-west-2, which sit in the Service Quotas console.
- Whether the Bedrock SLA's "APIs for models" wording covers ApplyGuardrail and InvokeGuardrailChecks.
- Whether InvokeGuardrailChecks calls are logged in CloudTrail like ApplyGuardrail.
- What the August and September 2026 Bedrock error-rate warnings covered. We read them only through StatusGator, and they don't name a component.
- Whether data sent to ApplyGuardrail is retained, since the Bedrock data-retention page covers inference requests only.

## Weaknesses

- Per-policy billing, so four paid policies on one request cost four times, and no free tier
- Classic tier covers English, French and Spanish only, and Standard tier uses cross-Region inference that can move prompts within a geography
- Quota numbers are mostly in the Service Quotas console, with public figures only for two US regions
- The Bedrock SLA covers APIs for models and doesn't name Guardrails
- No Guardrails change announced since 23 June 2026

## 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.

- Call ApplyGuardrail twice, once with source INPUT before the model and once with source OUTPUT after, since the policies that apply differ
- Use InvokeGuardrailChecks when you only need content, prompt-attack or PII scores. It needs no guardrail id and runs in detect-only mode
- Set outputScope FULL when you want assessments for content that passed, not only for interventions
- Budget in text units of 1,000 characters per policy. A 5,000-character tool result is five units on every paid policy
- Retry ThrottlingException (429) and ServiceUnavailableException (503) with exponential backoff, but treat a 400 ServiceQuotaExceededException as a quota to raise

## What the review panel asked for

- Add a free tier
- Published quotas per Region
- Guardrails in the SLA
- Guardrails entries in the document history
- billing for failed calls
- published accuracy figures
- a retention statement
- Publish Guardrails quotas for every region
- Name Guardrails in the SLA
- Say which errors to retry on every operation page
- Publish quotas for every Region
- retention terms for ApplyGuardrail
- CloudTrail coverage for InvokeGuardrailChecks

## 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.