Aidbox

# Fix list: Aidbox

From Anchor Terminal's listing at https://www.anchorterminal.com/tools/aidbox, the October 2026 research run, assessed 14 October 2026. Grade B, 64.2 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 Aidbox: 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, 25 out of 100, up to 9.4 more on the total

Why it scored 25: Read with the hosted rubric, because Aidbox is licensed software with a price rather than free software. No x402, MPP or L402 (0). Plan prices are public, Base from $1,900 a month or $19,000 a year, with no per-call price (10). A free development licence and hosted sandbox come with an aidbox.app account, limited to data without PHI, and whether a card is asked for was not established (15 of 20). A person registers on the aidbox.app portal and creates a licence 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. Reliability, 59 out of 100, up to 8.2 more on the total

Why it scored 59: Read with the local-software lines, because production Aidbox runs on the customer's infrastructure from Health Samurai's Docker images and the hosted sandbox is for development only. Installs from the official image healthsamurai/aidboxone (7.8 million pulls on Docker Hub) with PostgreSQL requirements stated (20). Source and CI are closed, so no public test suite could be seen (0). The public Aidbox/Issues tracker has 31 open issues, most from 2023, with support handled mainly on a customer Zulip (12 of 25). YYMM.minor versions with dated release notes that mark breaking changes, though the scheme is not semver (12 of 15). Stable and LTS channels are declared, with LTS lines supported for two years (15). Aidbox has no rate limiting of its own, by its hardening page, which this checklist does not score but an agent should know.

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.

## 3. Transparency & trust, 45 out of 100, up to 4.8 more on the total

Made of editorial 51, provenance 39.

Why it scored 45: Closed source, and no terms of service or licence agreement was found on the website or in the docs. Licence types and their limits are described in the docs (5 of 30). The privacy policy was last updated on 10 June 2021 and covers website visitors. The FAQ says licence checks send only the JWT token. No DPA, retention periods or sub-processor list were found (10 of 30). LTS lines carry published end-of-life dates, and a deprecated section lists retired functions (16 of 20). The FAQ discloses usage statistics with a setting to turn them off (20 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):

- Domain age: health-samurai.io, no registry record we could read (0 of 15)
- Endpoint on the vendor's domain:  is not on health-samurai.io (0 of 15)
- Terms of service: not found (0 of 10)
- Privacy policy: read, states 7 of the 8 things a reader expects (9.3 of 10)
- Status page: not found (0 of 10)
- security.txt: not found (0 of 10)

## 4. Security & auth, 74 out of 100, up to 4.6 more on the total

Why it scored 74: Each instance runs its own OAuth 2.0 server with client credentials, authorisation code, token exchange and SMART backend services with private key JWT. Token lifetime is set per Client. Basic auth with a client secret is also allowed (26 of 30). AccessPolicy rules per client, user or operation, plus compartments, label-based and organisation-based access, allow read-only clients. The MCP endpoints are closed by default, though the docs show a public policy for testing. No confirmation for destructive tools (15 of 20). A Health Samurai article of 24 August 2026 discusses guardrails for agents acting on FHIR data, with no injection guidance in the MCP docs (6 of 15). AuditEvent recording and an audit event topic for delivery elsewhere (14 of 15). ISO 27001:2022 claimed, a reporting address, remediation deadlines, signed images and SBOMs. No security.txt and no bug bounty found (13 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, 77 out of 100, up to 3.7 more on the total

Why it scored 77: No OpenAPI document was found. Each instance publishes a FHIR CapabilityStatement and StructureDefinitions, which are machine-readable (15 of 25). llms.txt and Markdown copies of every docs page (10). Reference pages describe each operation and setting, and the MCP page gives a one-line description per tool (14 of 20). FHIR resources are validated against profiles with required fields and value sets, while the MCP tools take whole resources as free JSON objects (12 of 15). Docs carry HTTP examples, and errors come back as FHIR OperationOutcome (11 of 15). Dated monthly release notes back to 2024 (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, 79 out of 100, up to 3.4 more on the total

Why it scored 79: `_elements`, `_count`, `_total` and `_summary` size responses (22 of 25). Bundle paging links, `_page`, chained search and SQL on FHIR (20). Errors are OperationOutcome with FHIR issue codes, seen on a live 404 from aidbox.app (14 of 20). Conditional create with `If-None-Exist`, conditional update and patch give safe retries. No MCP annotations are documented (10 of 20). Few required parameters and official SDKs for TypeScript, Python and C# (13 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. Maintenance & community, 80 out of 100, up to 1.8 more on the total

Why it scored 80: Release 2609 published on 8 October 2026 (30). Monthly releases plus bugfix minors, 2607, 2608 and 2609 in the last 90 days (20). Closed product with public dated release notes and a customer Zulip and email support. The public issue tracker has few recent entries (11 of 15 on the closed-service scale). Official SDKs exist for TypeScript, Python and C#, and aidbox-python-sdk was updated on 3 October 2026 (12 of 15). Signed images and SBOMs per release, with Trivy scans before release (7 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.

- unchecked: Aidbox's terms of service or licence agreement. None was found at /legal or in the docs, and the privacy policy refers to Terms of Use that were not located
- unchecked: whether aidbox.app sign-up asks for a card
- unchecked: GitHub organisation repository details beyond the public page, because the GitHub API refused us for its rate limit
- unchecked: the domain registration date for health-samurai.io
- The lead described Aidbox as having an open-source core. Aidbox itself is proprietary, and only examples and SDKs are on GitHub under MIT

## Weaknesses

- No built-in rate limiting or brute-force protection, by the vendor's hardening page
- The MCP module is alpha, its tools take whole resources as free JSON, and no tool annotations are documented
- No terms of service or licence agreement was found on the website, and the privacy policy dates from 10 June 2021
- Closed source with no public CI and no OpenAPI document
- No public status page, security.txt or sub-processor list was found

## 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.

- Ask the operator for a Client with an AccessPolicy limited to the task. Don't rely on the public MCP policy, which the docs mark unsafe
- Use `If-None-Exist` on creates and conditional update on identifiers so retries don't duplicate records
- Use `_elements` and `_count` to keep responses small, and follow the Bundle's paging links
- Throttle yourself. Aidbox applies no rate limit, so a burst lands on the customer's database
- Don't load real patient data into an instance on a development licence

## 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.