Microsoft Entra Agent ID

# Fix list: Microsoft Entra Agent ID

From Anchor Terminal's listing at https://www.anchorterminal.com/tools/microsoft-entra-agent-id, the October 2026 research run, assessed 8 October 2026. Grade BB, 74.4 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 Microsoft Entra Agent ID: 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). Agent 365 is listed at $15 a user a month and Microsoft 365 E7 at $99, and the docs say Agent ID is available to all Entra customers, but no page gives a per-agent price or says plainly which parts are free (10). Entra ID Free comes with a Microsoft cloud subscription. We didn't establish whether a new tenant can be opened without a card, so this line gets half (10). A person creates the tenant, holds an Agent ID role and creates the blueprint before any agent can act (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, 71 out of 100, up to 4.7 more on the total

Why it scored 71: List calls take `$select`, `$top`, `$filter`, `$search` and `$count`, with a default and maximum page of 100 (20). OData paging and filtering work on the main collection, but ownedObjects, deletedItems and owners can't filter by agent type and need client-side filtering (17). Named error codes such as `AgentIdentity_CredentialsNotSupported` say what to change (16). No idempotency key for creates, and sequential creates need retries. Token requests are safe to repeat (8). Creating an identity needs three fields, but getting a token is a two-step exchange that Microsoft's docs call complex and error-prone by hand, with an in-process library only for .NET and a sidecar container for everything else (10).

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. Security & auth, 83 out of 100, up to 3 more on the total

Why it scored 83: OAuth 2.0 with scoped Graph permissions. Agent identities can't hold credentials, the blueprint signs in with a managed identity or certificate, and client secrets are allowed but warned against (30). Entra refuses high-privilege roles and permissions such as Application.ReadWrite.All for agents, and an identity can be disabled. Conditional Access for agents needs an Agent 365 licence, and nothing asks a person to approve a destructive call (16). The service returns tokens and directory objects, not untrusted content (10). Audit and sign-in logs carry agentType and blueprintId, kept seven days on Free and 30 on P1 or P2, while the FAQ says Graph activity logs don't separate agents (12). MSRC disclosure policy and an identity bounty of $750 to $100,000, but microsoft.com's security.txt expired on 23 September 2026 and we didn't read certification reports (15).

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.

## 4. Transparency & trust, 74 out of 100, up to 2.3 more on the total

Made of editorial 63, provenance 85.

Why it scored 74: Closed service under the Microsoft APIs terms of use, last updated October 2025, with MIT SDK source (15). The privacy statement was updated in September 2026, Entra publishes a data residency page and log retention periods, and we didn't read the DPA (20). Microsoft Graph gives at least 24 months' notice before retiring a GA API or version, while several Agent ID operations are still on /beta, which can change without notice (18). Entra documents where tenant data is stored by geography. We didn't read the sub-processor list (10).

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 5 of the 7 things a reader expects, and has 3 clauses that cost points (2.3 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)

## 5. Schema & documentation, 87 out of 100, up to 2.1 more on the total

Why it scored 87: agentIdentity is in Microsoft's public Graph v1.0 OpenAPI file (25). learn.microsoft.com returns Markdown when asked with `Accept: text/markdown`, but learn.microsoft.com/llms.txt returned 404 (8). Concept, design-pattern and best-practice pages say when to use each flow and when to pick the sidecar over the .NET library (16). Graph resources are typed with required fields listed, though agentIdentity is an open type that inherits servicePrincipal properties that don't all apply (12). Request and response examples on each reference page and a list of 17 Agent ID error codes, without the HTTP status for each (11). v1.0 and beta versioning with a monthly dated Graph changelog and Entra 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.

## 6. Reliability, 91 out of 100, up to 1.8 more on the total

Why it scored 91: Microsoft's SLA page for Entra ID points to the Azure status history for incidents, and that page was readable with dated entries and reviews (20). The last 90 days show three Azure incidents, on 23 July, 29 September and 30 September 2026. None names Entra ID, the 23 July West US network fault lists Azure AD B2C, and Microsoft reports 99.999 per cent authentication availability for July, August and September. We scored between clean and minor because the page lists only wide incidents (25). Graph identity limits are published as numbers, such as 3,000 writes per 2 minutes 30 seconds per app and tenant (15). 429 carries Retry-After and the FAQ asks for exponential backoff, but there is no idempotency key and back-to-back creates can fail with 400 during replication (11). An SLA for Microsoft Entra ID is published and attainment is reported monthly. The SLA document itself didn't load for us (10). Agent ID has been generally available since April 2026 and the agentIdentity API is on Graph v1.0 (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.

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

Why it scored 80: Microsoft.Identity.Web 4.16.0, which carries the agent identity package and the sidecar, was tagged on 30 September 2026, and the Agent ID FAQ was updated on 1 October (30). Eight tags from 4.13.0 on 9 July to 4.16.0, and Graph changelog entries for agent identities in August and September (20). A closed service with public release notes and a public SDK tracker showing 340 open issues. We didn't read reply times (10). The current in-process SDK is .NET only, with the sidecar container for other languages (12). Azure Pipelines and CodeQL are configured, and 4.14.x raised dependencies for four CVEs (8).

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.

- unchecked: the SLA document for Microsoft Entra ID. The link redirected to a general SLA index and we didn't read the availability figure or which licences it covers.
- unchecked: whether a new Entra tenant can be created without a payment card.
- Whether creating and using agent identities on Entra ID Free carries any charge. The docs say Agent ID is available to all Entra customers and give no per-agent price.
- unchecked: the Microsoft Products and Services DPA, the sub-processor list and the Product Terms for Entra and Agent 365.
- unchecked: certification reports (SOC 2, ISO 27001) for Entra ID, and reply times on the microsoft-identity-web issue tracker.
- unchecked: status.cloud.microsoft, which needs JavaScript. The incident record rests on the Azure status history and Microsoft's own monthly SLA table.
- The Agent ID how-to page for creating identities still shows the /beta Graph endpoint while the reference documents the same call on v1.0.
- The logs page says audit events carry agentType, while the FAQ says audit logs don't distinguish agent identities by default. We didn't test which is current.

## Weaknesses

- Conditional Access, ID Protection and governance for agents need Microsoft Agent 365, listed at $15 a user a month on yearly billing
- Microsoft's docs call hand-written token exchange complex and error-prone and steer developers to a .NET library or a sidecar container
- Creating a blueprint, a principal and an identity in quick succession can fail with 400 until the directory replicates
- Audit and sign-in logs are kept seven days on Entra ID Free and 30 days on P1 or P2
- microsoft.com's security.txt passed its Expires date on 23 September 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.

- Request tokens in two steps. The blueprint gets an exchange token with fmi_path set to the agent identity's client ID, then the agent identity trades it for a resource token
- Retry with exponential backoff when a create returns `400 Object with id not found` straight after creating its parent object
- Send displayName, agentIdentityBlueprintId and a sponsor reference when creating an agent identity. All three are required
- Don't use the interactive `/authorize` flow. Agent identities are confidential clients and can't sign in to a page
- Keep the sidecar off any public network. Its `/AuthorizationHeader` endpoint hands out tokens to whoever can reach it

## 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.