Shufti

# Fix list: Shufti

From Anchor Terminal's listing at https://www.anchorterminal.com/tools/shufti, the October 2026 research run, assessed 9 October 2026. Grade C, 57.8 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 Shufti: 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, 47 out of 100, up to 8.6 more on the total

Why it scored 47: REST responses can't be trimmed by field, and AML and e-IDV responses are long. The MCP server has 25 tools, a knowledge-only scope, and status answers built from a fixed list of fields (12). No endpoint lists verifications, so every read needs a stored reference. Paging was found only on Travel Rule transactions (6). Errors name the service, the key and a message, with HTTP codes and decline codes documented (15). No idempotency key. The caller's `reference` identifies a request and a 409 is listed, but the docs do not say a repeated reference is refused. The MCP delete tool asks for explicit confirmation, and tool annotations were not read (7). A request needs only a reference and one service object, and the MCP server fills defaults and reports them in `defaults_applied`. No server-side SDK in any language (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, 35 out of 100, up to 8.1 more on the total

Why it scored 35: No x402, MPP or L402 (0). The pricing page gives $0 for Free Forever and a starting price of $1.50 a verification for Essentials, with the rate varying by volume and service and Enterprise quoted. We scored 15 of 20 because only a starting price is public (15). Free Forever covers 10 verifications a month with no card (20). Signup is a browser flow and keys come from the back office. The MCP server's Explore Shufti mode needs no account but reaches only the knowledge tools (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. Reliability, 65 out of 100, up to 7 more on the total

Why it scored 65: Graded on the REST API with the hosted lines. A status page at status.shuftipro.com, hosted by UptimeRobot, with 90-day uptime and an update history (20). The page draws its data by script from a path its robots.txt disallows, so no incident record was read and the line takes the no-readable-history score (5). Rate limits are published at 60 requests a minute per IP address for production accounts and 20 for trial accounts (15). A 429 is in the HTTP code table. No Retry-After header, backoff guidance or idempotency key was found. Failed callbacks retry automatically and can be resent by API once an hour (5). Schedule 1 of the public terms targets 99 per cent monthly uptime on a reasonable endeavours basis, with no service credits found (10). The REST API is 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. Schema & documentation, 58 out of 100, up to 6.8 more on the total

Why it scored 58: No OpenAPI file or other machine-readable contract was found. Postman collections are linked from the endpoint pages, and the MCP tool schemas sit behind OAuth and were not read (5). llms.txt, a 2.9 MB llms-full.txt and a Markdown twin of every page (10). Each service has a how-it-works page and parameter tables that state purpose, and the MCP tool list gives one line a tool (12). Parameter tables give type, length limits, accepted values and required flags, though most flags are strings holding 0 or 1 and every service shares one endpoint (10). Samples in nine languages, an HTTP code table, an events table, sample responses per event and decline codes per service (13). The API has no version in its URL or headers. The revision history is dated and frequent, and its published page contains unresolved merge conflict markers (8).

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.

## 5. Security & auth, 62 out of 100, up to 6.7 more on the total

Why it scored 62: The REST API uses one Client ID and Secret Key pair as Basic auth, with no scopes. The secret is shown once and replaced by generating a new one, and a one-hour access token is an alternative. The MCP server adds OAuth 2.1 with PKCE, three scopes, 24-hour access tokens and rotating refresh tokens, though its login page takes the same key pair and stores the secret encrypted (22). MCP scopes are enforced when tools are listed and on every call, the delete tool needs confirmation, callback domains must be registered, and the back office has five preset roles plus custom ones. No read-only REST key (12). MCP status answers leave out extracted document data and the provider payload by design. REST responses carry text read from documents and watchlist hits, with no injection guidance found (7). Callback logs, a login history and an activity log for screening cases. MCP usage records are kept by Shufti and not shown to the operator (9). SOC 2 Type II, ISO 27001:2022, PCI DSS and Cyber Essentials Plus per the certifications page. security.txt returns 404, and no disclosure policy or bug bounty was found (12).

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.

## 6. Maintenance & community, 72 out of 100, up to 2.5 more on the total

Why it scored 72: The revision history's newest entry is dated 6 October 2026 (30). It has 19 dated entries between 24 July and 6 October 2026 (20). Closed service with a public revision history, a tech support address and a help desk described in the terms. No public developer forum was found (9). No server appears in the official MCP registry and there is no server-side SDK. The mobile SDKs are current, with Android 3.0.8 dated 31 August 2026 (8). The npm package `shuftipro-onsite-mobilesdk` is at 1.1.2 with 673 downloads in the week to 4 October 2026. CI was not checked (5).

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.

## 7. Transparency & trust, 71 out of 100, up to 2.5 more on the total

Made of editorial 51, provenance 90.

Why it scored 71: Closed service with public terms, version 11.1 with effect from 13 May 2026, and every earlier version linked (15). The Services Privacy Notice, version 1.2 of September 2026, sets out processor and controller roles and a retention schedule, and the terms carry a data processing schedule. The terms keep personal data for seven years after the services end where the client gives no instruction, and both documents say Shufti may act as independent controller for model training. The notice's United States section gives twelve months as the default for other verification data, which differs from the terms (22). Deprecations appear in the revision history with dates. Standard KYB was deprecated on 6 October 2026 with its pages removed the same day and no notice period, while the terms promise ten business days' written notice of non-security updates (8). Sub-processors are described by category and the list is given to clients on request. Webhook IP addresses are published for Europe and the United States (6).

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):

- security.txt: not found (0 of 10)

## 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 incident history on status.shuftipro.com, which loads from /api/, a path its robots.txt disallows
- unchecked: the Vanta trust centre at trust.shuftipro.com, which is drawn by script, so the certification reports and any sub-processor list there were not read
- unchecked: the MCP tool definitions, annotations and input schemas, which sit behind OAuth. We did not connect to the server
- unchecked: GitHub stars and the licences of the mobile SDK repositories. The GitHub API refused us for its rate limit
- Whether clients were told in advance of the 14 September 2026 change that replaced decline code SPDR398 with SPDR370 for United States driving licences. The revision history entry carries the same date as the change, and no deduction was taken
- Whether a repeated `reference` is refused. A 409 is listed without saying when it is returned
- The Get Started page says an access token is valid for one hour and can be used once only, and the endpoint page says it applies for one hour. We did not test which holds
- Whether the Free Forever plan asks for no card at signup. The pricing page says so and we did not sign up
- The default retention period. The terms say seven years after the services end and the privacy notice's United States section says twelve months
- The lead named shufti.ai as the vendor URL. It is a parked domain for sale, and the vendor's site is shuftipro.com. The lead also did not mention the MCP server
- npmWeekly is `shuftipro-onsite-mobilesdk` (28 September to 4 October 2026), a React Native capture SDK, since no server-side package exists

## Weaknesses

- No OpenAPI file or other machine-readable contract was found. The reference is prose tables plus Postman collections
- The REST credential is one Client ID and Secret Key pair with no scopes, sent as Basic auth on every call
- No idempotency key, Retry-After header or backoff guidance was found for the documented 429
- No server-side SDK. Official packages cover Android, iOS, Flutter, React Native and Cordova capture only
- Standard KYB was deprecated on 6 October 2026 and its pages removed the same day, with no notice period stated
- The sub-processor list is available to clients on request only, and security.txt returns 404

## 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.

- POST every verification to `https://api.shuftipro.com/` with a unique `reference` of 6 to 250 characters and one object per service. Read results from `/status` with that reference
- Register the callback domain in the back office first. An unregistered `callback_url` is rejected
- Stay under 60 requests a minute per IP on a production account and 20 on a trial account
- Check the `Signature` response header. Accounts created after 15 March 2023 hash the Secret Key with SHA-256 before appending it to the raw response
- Through MCP, identity checks return a `verification_url` for the person to open. No tool accepts an image, so use the REST API for offsite proofs

## 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.