Azure Key Vault

# Fix list: Azure Key Vault

From Anchor Terminal's listing at https://www.anchorterminal.com/tools/azure-key-vault, the October 2026 research run, assessed 8 October 2026. Grade BB, 74.7 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 Azure Key Vault: 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). $0.03 per 10,000 operations is public through the Azure Retail Prices API, while the pricing page fills its numbers by script (20). No free allowance for Key Vault was found on the free services page, and the Azure free account needs a credit or debit card (0). A person signs up in a browser with a phone number and a card (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: API reading. Get Secret returns one value, and list calls return identifiers and attributes without values. There is no field selection (20 of 25). `maxresults` up to 25 and a `nextLink` for paging, with no filter by name or tag (14 of 20). Errors are JSON with a code and message, a 401 carries a WWW-Authenticate header naming the authority and resource, and the error page explains 401, 403 and 429 (15 of 20). Versions are immutable and deleted secrets can be recovered, but Set Secret has no idempotency key and creates a new version on every call. Key Vault has no MCP server of its own to annotate. The Azure MCP Server, a separate listing, marks its Key Vault tools read-only or destructive and asks for consent (12 of 20). SDKs for Python, JavaScript, .NET, Java and Go, and a read needs only the vault and secret name. `api-version` is required on every REST call (14 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. Reliability, 87 out of 100, up to 2.6 more on the total

Why it scored 87: Hosted reading. Azure status page with a history of post-incident reviews (20). The history lists three incidents in the last 90 days, West US network connectivity on 23 July 2026 (14:44 to 19:41 UTC), Azure OpenAI and Cognitive Services in Sweden Central on 29 September, and gateway services in several regions on 30 September. None names Key Vault. The page lists only broad incidents and the West US fault blocked traffic entering or leaving that region, so we count the record as minor (20 of 30). Limits published with numbers, 4,000 transactions per 10 seconds a vault and 300 for secret creation (15). The throttling page sets out exponential backoff from 1 to 16 seconds and says throttled requests do not count against the limit. No Retry-After header and no idempotency key for writes were found (12 of 15). The SLA document of 1 October 2026 commits to 99.99 per cent with 10 and 25 per cent credits (10). Generally available (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. Security & auth, 86 out of 100, up to 2.5 more on the total

Why it scored 86: Microsoft Entra ID OAuth 2.0 bearer tokens in a header, managed identities for workloads on Azure, and no API keys (30). Key Vault Secrets User reads secret contents only, Key Vault Reader sees metadata only, roles can be assigned on one secret, purging needs its own role, and deleted secrets stay recoverable for 7 to 90 days. Nothing asks for approval on a write (18 of 20). The service returns no untrusted content (10). AuditEvent logs record each operation with caller identity and IP address, but only after the owner adds a diagnostic setting (12 of 15). MSRC coordinated disclosure and bounty policies and a SOC 2 Type 2 report covering Azure. The microsoft.com security.txt passed its Expires date on 23 September 2026, and we could not confirm Key Vault on the audit scope list (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.

## 5. Schema & documentation, 85 out of 100, up to 2.4 more on the total

Why it scored 85: A public OpenAPI 2.0 document for the secrets data plane, version 2025-07-01, with 12 operations (25). learn.microsoft.com/llms.txt returns 404, but every Learn page we asked for with `Accept: text/markdown` came back as Markdown, so half (5 of 10). Each operation states its purpose and the permission it needs, and the throttling page says when not to use Key Vault, such as for customers' secrets at high throughput (16 of 20). A pattern on secret names, a 1 to 25 range on `maxresults` and an enum for the recovery level. The value is a free string and tags a free object (13 of 15). Sample requests and responses on each reference page. Errors appear only as a generic KeyVaultError, with a separate page for 401, 403 and 429 (11 of 15). Dated `api-version` values and a what's new page, whose newest entry is March 2026 (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. Transparency & trust, 77 out of 100, up to 2 more on the total

Made of editorial 67, provenance 86.

Why it scored 77: Closed service under Microsoft's Product Terms, with MIT client libraries (15 of 30). The privacy statement, last updated September 2026, sends Azure customer data to the Product Terms and the Data Protection Addendum. The Key Vault docs say Microsoft doesn't see or extract your data and that logs stay in storage the customer owns. We did not read the addendum itself (22 of 30). Microsoft's Modern Lifecycle Policy promises 12 months' notice before support ends, and Key Vault gives a dated retirement, 27 February 2027, for control plane API versions before 2026-02-01 (18 of 20). Replication stays within the region and its paired region in the same geography. The sub-processor list sits on the Service Trust Portal, a script-drawn page we could not read (12 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):

- Terms of service: read, states 4 of the 7 things a reader expects, and has 2 clauses that cost points (3.4 of 10)
- Privacy policy: read, states 8 of the 8 things a reader expects, and has 1 clause that costs points (8 of 10)
- security.txt: published but past its Expires date (5 of 10)

## 7. Maintenance & community, 80 out of 100, up to 1.8 more on the total

Why it scored 80: Read as a closed service with official SDKs. azure-keyvault-secrets 4.11.3 reached PyPI on 2 October 2026 (30). Three Python releases in the last 90 days, 4.11.1 on 12 August, 4.11.2 on 25 August and 4.11.3 on 2 October (20). A public what's new page whose newest entry is March 2026, Microsoft Q&A and the SDK issue trackers on GitHub, which we did not read (10 of 15). Official SDKs current in Python, JavaScript, .NET, Go and Java (15). The SDKs are released from the azure-sdk monorepos, CI not checked (5 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.

## 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 lead was right on the interface and the docs. Its URL redirects to https://azure.microsoft.com/en-us/products/key-vault.
- unchecked: whether Key Vault is on Azure's SOC 2 audit scope list. The scope page we loaded did not name services, and the reports sit on the Service Trust Portal.
- unchecked: the Microsoft sub-processor list and the Data Protection Addendum. The Service Trust Portal page is drawn by script and the addendum page did not load.
- unchecked: the Azure bounty programme page, which answered 403. The bounty policy link comes from security.txt.
- unchecked: incidents limited to one region or one service, since the public status history lists only broad incidents.
- unchecked: CI and open issues for the azure-sdk repositories.
- The Python changelog on GitHub still marks 4.11.3 as unreleased, while PyPI shows it uploaded on 2 October 2026. lastRelease follows PyPI.
- Not established: whether 429 responses carry a Retry-After header, and whether throttled calls are billed.
- The Product Terms page showed no effective date when read.

## Weaknesses

- No managed rotation for secrets. Key Vault raises a near-expiry event 30 days ahead and the owner's Azure Function does the rotation
- Audit logging is off until a diagnostic setting sends AuditEvent logs to a storage account, event hub or Azure Monitor
- Set Secret has no idempotency key, so a retried write adds another version
- Listing returns at most 25 secrets a page with no name or tag filter
- No free allowance for Key Vault was found, and an Azure account needs a phone number and a credit or debit card

## 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.

- Request a token for the resource https://vault.azure.net and send it as a Bearer header. Every call must carry `api-version`, such as 2025-07-01, because there is no default
- Ask for the Key Vault Secrets User role on the vault or the single secret. Key Vault Reader returns metadata only, not values
- Cache secret values in memory. A vault allows 4,000 reads and 300 secret creations per 10 seconds, and a subscription five times that
- On 429 wait 1, 2, 4, 8 then 16 seconds. Throttled requests do not count against the limit
- Don't retry Set Secret blindly. Each successful call creates a new version of the secret

## 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.