Backblaze B2 by Backblaze

# Fix list: Backblaze B2

From Anchor Terminal's listing at https://www.anchorterminal.com/tools/backblaze-b2, the October 2026 research run, assessed 1 October 2026. Grade BB, 75.4 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 Backblaze B2: 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, 60 out of 100, up to 8 more on the total

Why it scored 60: Status page at status.backblaze.com (20). It renders only with JavaScript and has no Statuspage feed, so we couldn't read its history (5). No rate limits with numbers. The native API docs say only that B2 may throttle requests per account (0). The docs list which errors to retry (401 expired_auth_token, 408, 429, 500, 503), advise exponential backoff on 503 and a fresh upload URL after a failed upload, and the MCP server retries 408, 429 and 5xx itself (15). SLA of 99.9 per cent monthly uptime for all B2 customers, with 5 or 10 per cent credits (10). GA (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.

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

Why it scored 40: No x402, MPP or L402 (0). Per-TB storage and per-call prices public without a login (20). The first 10 GB are free and the sign-up page says no credit card is required (20). A person signs up in a browser. The Partner API can create accounts, but only for partners holding a master key (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.

## 3. Agent ergonomics, 85 out of 100, up to 2.4 more on the total

Why it scored 85: The full MCP surface is 40 tools with 49,500 characters of input schema before descriptions (5). Registration follows the key's capabilities, so a read-only key sees 20 tools and 15,400 characters, and a non-master key 37 (10 back). ListObjectsV2 with prefix, delimiter, maxKeys and continuation tokens (20). Named, documented error codes from the API and the server (17). Every tool carries readOnlyHint, destructiveHint and idempotentHint (17 read-only, 15 destructive in the full set), and key-minting tools take idempotency keys (20). Official SDKs in Python and Java plus the TypeScript MCP package, and few required parameters (13).

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.

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

Made of editorial 72, provenance 75.

Why it scored 74: Closed service with dated terms (2026-04-16), and the MCP server is MIT (20, five over the closed-service line for the open server). Privacy policy and terms per the 30 September check, and the MCP server's PRIVACY.md says the publisher receives no credentials, object data or telemetry. We didn't read a DPA this run (18). A stated deprecation policy, at least a year's notice before any native API version is dropped (20). The MCP server never phones home. Data regions are chosen per account, but we didn't read a sub-processor list (14).

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):

- Endpoint on the vendor's domain: s3.us-west-004.backblazeb2.com is not on backblaze.com (0 of 15)
- security.txt: not found (0 of 10)

## 5. Schema & documentation, 89 out of 100, up to 1.8 more on the total

Why it scored 89: No OpenAPI for the B2 APIs, but every MCP tool has a typed JSON Schema input, checked against committed contract fixtures (25). No llms.txt at www.backblaze.com/docs (404). The MCP repo ships AGENTS.md and a Markdown skills pack for agents (5). Tool descriptions state purpose and point elsewhere when a tool is the wrong one, s3_put_object for example sends anything over 1 MiB to presigned URLs or multipart (18). Typed inputs with enums and bounds (maxKeys 1 to 1,000, expiresIn up to 604,800 seconds); custom metadata is the one free-form map (13). The native API documents a JSON error shape and per-call error codes, and the MCP server returns named errors (13). Native API versions v1 to v4 dated on one page, and a Keep a Changelog file with semver for the MCP server (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. Security & auth, 90 out of 100, up to 1.8 more on the total

Why it scored 90: Application keys scoped to one or more buckets, a name prefix and named capabilities, with an optional expiry, and revocable. AssumeRole temporary credentials arrived on 30 September 2026 for Enterprise customers only (30). Read-only keys, and the MCP server's destructive gate (confirm on stdio, block on HTTP, with MCP elicitation), plus Object Lock against deletion (20). The MCP server keeps object bytes out of the model by default through presigned URLs and saveToPath, but we found no prompt-injection guidance (10). Bucket Access Logs on the service and a values-redacted audit log in the MCP server (13). SOC 2 Type 2, a public Bugcrowd bug bounty and a SECURITY.md in the MCP repo, but no security.txt on backblaze.com (17).

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.

## 7. Maintenance & community, 90 out of 100, up to 0.9 more on the total

Why it scored 90: MCP server v0.2.2 on 29 September 2026 and the first STS production wave on 30 September (30). Six MCP releases since 18 August (0.1.0 to 0.2.2) (20). The MCP repo merges pull requests daily (numbered past #450), but we couldn't see issue replies from git, and the B2 release notes page stopped in 2016 (15). In the official MCP registry as io.github.backblaze-labs/b2-mcp, per the 30 September check, and current official SDKs (15). CI for tests, contract checks, CodeQL and mutation testing, and transitive advisories patched in 0.2.2 (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.

- unchecked: B2 incident history for the last 90 days, since status.backblaze.com renders only with JavaScript
- unchecked: issue response times on backblaze-labs/b2-mcp, which aren't visible from git
- Whether the IAM and STS API will reach accounts outside the Enterprise Web Console, and when; the page names waves on 2026-09-30 and 2026-11-05 only for Limited Availability
- Backblaze doesn't publish numeric rate limits; we don't know where throttling starts

## Weaknesses

- No rate limits published with numbers; the docs say only that B2 may throttle per account
- status.backblaze.com renders only with JavaScript, so its incident history can't be read by a script
- STS temporary credentials are in Limited Availability for Enterprise customers only, from 30 September 2026
- The full MCP surface is 40 tools and 49,500 characters of input schema before descriptions
- No llms.txt or OpenAPI for the B2 APIs, and the B2 release notes page stopped in 2016

## 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.

- Take the region from the key's S3 endpoint (the second label of s3.<region>.backblazeb2.com) and pass it as the SDK region
- Mint a per-task application key for one bucket and a name prefix, with an expiry, rather than holding the master key
- Move bytes with s3_get_presigned_url so file contents never pass through the model
- On 401 expired_auth_token call b2_authorize_account again; after a failed upload fetch a new upload URL; on 503 back off exponentially
- Expect fewer than 40 tools with a non-master or read-only key; that's the server trimming tools to the key

## What the review panel asked for

- Programmatic account signup
- Numeric rate limits
- Statuspage feed
- a current service changelog
- Ship a smaller core profile
- Publish OpenAPI for the native API
- numeric rate limits
- llms.txt
- Publish numeric rate limits
- Offer a status feed
- State failed-call billing
- STS for every account
- a security.txt

## 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.