# Fix list: Didit

From Anchor Terminal's listing at https://www.anchorterminal.com/tools/didit, the October 2026 research run, assessed 8 October 2026. Grade BB, 75 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 Didit: 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. Agent ergonomics, 69 out of 100, up to 5 more on the total

Why it scored 69: Graded on the REST API. No field selection, and the decision payload returns every feature result as an array. Scoped keys can null out media links (12). `GET /v3/sessions/` takes `limit` and `offset`, returns `next` and `previous` links and filters on status, workflow, dates, country and `vendor_data` (18). Errors return `detail` or a field-keyed object, with two shapes on some routes and a bare array on password rules, each documented (14). `Idempotency-Key` on AML screening, transactions and wallet calls, and session creation returns the existing unfinished session for the same `vendor_data`. The MCP server marks read and destructive tools and asks for `confirm: true` on deletes (18). `workflow_id` is the only required field for a session. No server-side SDK was found in any language, only capture SDKs (7).

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.

## 2. Payments & pricing, 60 out of 100, up to 5 more on the total

Why it scored 60: No x402, MPP or L402 (0). Per-check prices are public without a login, such as $0.15 for ID verification and $0.20 for AML screening (20). 500 free checks a month per core feature with no card, changing to $10 of monthly credit on 1 November 2026 (20). An agent can register and receive an API key through `POST /auth/v2/programmatic/register/` and `verify-email` with no browser. It needs a real inbox for the emailed code, and we scored the line in full (20).

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. Security & auth, 76 out of 100, up to 4.2 more on the total

Why it scored 76: Named API keys take read or write access per resource, workflow and status limits, an IP allowlist, an expiry date, rotation with a 24-hour overlap and instant revoke. Keys travel in a header only. The primary key has full access, and the key returned by registration is that key (28). Read-only key presets, keys limited to approved sessions, `confirm: true` on MCP deletes and OAuth scopes that split management from verification. The MCP server's role check defaults to a mode that lists every tool and leaves refusal to the backend (16). Responses carry text read from documents, AML and adverse media hits. No guidance on injected content was found in the docs or the MCP repository (3). Audit logs record each API call with key name, path, status and IP, kept for 365 days (15). SOC 2 Type 2 issued 30 July 2026, ISO/IEC 27001:2022, yearly external penetration tests and a reporting address. No security.txt, no bounty terms and no public advisories found (14).

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

Why it scored 80: Graded on the REST API v3 with the hosted lines. Status page on incident.io at status.didit.me with Core APIs, Business Console and Hosted Verification Web App components (20). Three incidents in the last 90 days, all marked major by Didit and all partial outages of the core APIs traced to the primary database. 22 July 2026 lasted 21 minutes, 15 August about 65 minutes and 4 September about 40 minutes. The page reports 99.93 per cent uptime for Core APIs from July to October. We scored 10, between the lines for one major outage and several, because each was partial and about an hour or less (10). Limits are published per scope, 600 GET and 300 write requests a minute per key (15). A 429 carries `Retry-After` and `X-RateLimit-*` headers, backoff guidance is written out, and `Idempotency-Key` is documented on 31 operations (15). A public SLA commits to 99.9 per cent monthly uptime with service credits, excluding free-tier use (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.

## 5. Transparency & trust, 63 out of 100, up to 3.2 more on the total

Made of editorial 53, provenance 73.

Why it scored 63: Closed service with public, dated and versioned terms, and an MIT licence on the MCP server (17). A privacy policy, a data processing addendum inside the Business Terms and retention settings from 30 days to 10 years, unlimited by default. The terms allow model training on verification data by default with a console opt-out. The addendum says the sub-processor list is published at `/terms/sub-processors`, while that address shows the legal index, which says the list is sent after a signed NDA (20). `X-Signature-Simple` is marked deprecated with no end date, the terms promise 30 days' notice of changes to terms and prices, and the pricing change of 1 November was announced on 2 October. No API deprecation policy was found (8). Data is processed in the EU on AWS by default with an Australian data plane. The sub-processors aren't named in public (8).

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: didit.me, no registry record we could read (0 of 15)
- Terms of service: read, states 7 of the 7 things a reader expects, and has 1 clause that costs points (8 of 10)
- security.txt: not found (0 of 10)

## 6. Schema & documentation, 88 out of 100, up to 2 more on the total

Why it scored 88: A public OpenAPI 3.0.0 spec with 206 paths and 256 operations, and a second spec for the account routes (25). llms.txt and a Markdown copy of each docs page (10). 215 of the 256 operations have descriptions longer than 80 characters, covering billing, idempotency and which key access is needed. Few say when not to use an endpoint (17). 586 enum declarations, required lists and limits. Error bodies are loosely typed, and session creation returns either an array of strings or a plain string per field (12). 184 operations carry examples and status codes from 400 to 503 are described (13). A `/v3` path and a monthly changelog. Entries are grouped by month, with no day given, and the session routes still list v1 and v2 paths (12). The 88 total takes 1 off for the spec's 3.2 MB size, which no agent can load whole.

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.

## 7. Maintenance & community, 82 out of 100, up to 1.6 more on the total

Why it scored 82: The MCP server was released as v5.1.34 on 8 October 2026 and the September 2026 changelog covers the API (30). Monthly changelogs for July, August and September and three MCP tags in the period (20). Closed service with a public changelog and support by email, chat and WhatsApp. No public forum was found, and the MCP repository has no stars or open issues to judge replies by (10). The MCP server is in the official registry as `me.didit/mcp`, though the entry is at 5.0.1, and the capture SDKs are current (15). The MCP repository has a CI workflow that builds and tests, and publishes to npm from release tags. Whether CI passes wasn't checked (7).

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 registration date of didit.me (whois returned nothing and rdap.org has no service for .me)
- unchecked: whether CI passes on the default branch of didit-protocol/mcp, and the pages of the Business Console, which need a login
- Whether the August 2026 change to session deletion responses (204 to 200 with a body) was announced before it shipped. The changelog labels it breaking and we found no earlier notice, so no deduction was taken.
- The SLA names `api.didit.me` as the covered API while the documented base URL is `verification.didit.me`. Whether the SLA covers that host wasn't established.
- The sub-processor list. The addendum says it is published at `/terms/sub-processors`, the legal index says it is sent after a signed NDA, and the privacy policy says on request.
- Whether the limit of 600 session creations a minute applies to free accounts in practice. The docs say a defined 10 a minute free-tier limit is not applied.
- The lead was right on the interface and the 500 free checks, which end on 1 November 2026. The SDKs are capture clients, with no server-side library.

## Weaknesses

- Three incidents marked major on status.didit.me from 22 July to 4 September 2026, each a partial outage of the core APIs traced to the primary database
- No server-side SDK in any language. The published SDKs are capture clients for web, iOS, Android, React Native and Flutter
- The data processing addendum says the sub-processor list is published at `/terms/sub-processors`, but that address shows the legal index, which says the list is sent after a signed NDA
- Verification data is used for model training by default until an organisation owner turns it off in the console
- The hosted MCP server lists 156 tools to a signed-in user, 23 of them destructive, and role checks run in a mode that logs without hiding tools by default

## 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.

- Register with `POST https://apx.didit.me/auth/v2/programmatic/register/`, then `verify-email` with the emailed 6-character code. Use a real inbox, because reserved test domains return 500.
- Send the key as `x-api-key` to `https://verification.didit.me/v3/`. The JWT from registration works only on `apx.didit.me`.
- Create a workflow before `POST /v3/session/`. `workflow_id` is the only required field, and an unfinished session with the same `vendor_data` is returned again.
- Read results from webhooks and use `GET /v3/session/{sessionId}/decision/` for back-fill. Every per-feature result is a plural array.
- The MCP server at `https://mcp.didit.me/mcp` takes OAuth sign-in only, never an API key. Approve `didit:verification` alone when the task doesn't change workflows or keys.

## 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.