Plaid by Plaid

# Fix list: Plaid

From Anchor Terminal's listing at https://www.anchorterminal.com/tools/plaid, the October 2026 research run, assessed 1 October 2026. Grade BB, 70 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 Plaid: 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, 15 out of 100, up to 10.6 more on the total

Why it scored 15: No x402, MPP or L402 (0). No prices without a login. The pricing page names Pay as you go, Growth and Custom and still says the first 200 calls are free, with rates shown only after a Production application (0). Sandbox is free, needs no card and doesn't expire; the Trial plan's 10 free Production Items in the US and Canada is per the 30 September check, and we didn't confirm whether it asks for a card (15 of 20). A person signs up in a browser and applies for Production (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, 68 out of 100, up to 6.4 more on the total

Why it scored 68: Statuspage at status.plaid.com with per-product and per-bank components and history back to 5 June 2026 (20). From 3 July to 1 October Plaid's own API had minor incidents only, the longest elevated API errors on 24 August for about 15.5 hours. Six incidents were marked major, five of them single banks (Bank of America for about 22 hours from 2 August and 1.5 hours on 26 September, Wells Fargo for about 3 hours on 10 July) and one the Dashboard login on 16 July. We put that at 15 of 30, between the minor-only and one-major bands, because no major hit Plaid's core API. Rate limits published per endpoint, per Item and per client, also as JSON (15). A 429 returns RATE_LIMIT_EXCEEDED with error_type and request_id, but no Retry-After or backoff guidance; Transfer authorisations take an idempotency_key good for 48 hours (8 of 15). No SLA found (0). Core API 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.

## 3. Security & auth, 67 out of 100, up to 5.8 more on the total

Why it scored 67: client_id and secret travel in the JSON body or as headers, never in a URL, with separate secrets per environment. Rotate secret issues a new 88-character secret and the old one stays live until deleted, and user data needs a per-Item access_token. No scopes on the team secret (22 of 30). Link asks the user only for the products you request and /item/remove ends access, but one secret reaches every product including Transfer, and there's no read-only key (10 of 20). Returns bank records with merchant-written descriptions; no guidance on treating them as untrusted (7 of 15). Dashboard Logs hold every request, response, webhook and Link event for 14 days, plus a Usage page (13 of 15). security.txt valid to 31 December 2026 with a HackerOne programme; SOC 2 or ISO 27001 not stated on the pages we read (15 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.

## 4. Agent ergonomics, 82 out of 100, up to 2.9 more on the total

Why it scored 82: Response size controls include the /transactions/sync cursor with count, count and offset on /transactions/get, and account_ids filters (18 of 25). Cursor and offset pagination with filters (20). Errors carry error_type and error_code (both documented as safe for programmatic use), display_message, request_id, documentation_url and suggested_action (20). Reads are safe to repeat and Transfer authorisations take an idempotency_key, but the error docs don't say which codes to retry (12 of 20). Official SDKs released together, plaid-node 47.0.0 and plaid-python 44.0.0 on 1 September 2026; every call needs client_id and secret, and a real Item needs the Link front end (12 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.

## 5. Transparency & trust, 81 out of 100, up to 1.7 more on the total

Made of editorial 61, provenance 100.

Why it scored 81: Closed service with published developer policy and end-user agreements per region; SDKs and the sandbox MCP server are MIT (15). End User Privacy Policy updated 8 December 2025 says data is kept only as long as needed, with listed exceptions and no periods; no DPA or subprocessor list on the legal index (15 of 30). Dated deprecations in the changelog, such as account subtype changes on 11 October 2026 and the Cash Flow Updates migration deadline of 20 August 2027 (20). Data from the UK and EEA is transferred to the US and stored in AWS regions; no named subprocessor list found (8 of 20). Regulatory standing counts here as an addition to the checklist (+5 for a named regulator with a register number, +3 for a named regulator alone). The privacy policy says Plaid Financial Ltd answers to the FCA and Plaid B.V. to De Nederlandsche Bank, with no firm reference numbers on the page, and the US has no AISP licence to hold (+3).

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.

## 6. Maintenance & community, 83 out of 100, up to 1.5 more on the total

Why it scored 83: plaid-node 47.0.0 on 1 September 2026, and the changelog of 24 September lists new Link SDK releases (30). Four plaid-node releases since 23 July and nine dated changelog entries since 2 July (20). Closed service with a dated changelog and Dashboard support with case management since July 2026; GitHub issue replies weren't sampled because the GitHub API wasn't open to us (12 of 15). Official SDKs regenerated from the OpenAPI file at each release (15). plaid-node has no test workflow in .github, while the MCP repo runs pytest on pull requests (6 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.

## 7. Schema & documentation, 93 out of 100, up to 1.1 more on the total

Why it scored 93: Public OpenAPI file in plaid/plaid-openapi, at 2020-09-14_1.740.1 per the plaid-node 47.0.0 changelog (25). llms.txt with about 250 entries and a Markdown twin of each docs page (10). Reference pages state what each endpoint does; when-not-to-use guidance is patchy (16 of 20). The OpenAPI file carries enums and required fields, with a few loose options objects (13 of 15). Each error code has its own docs page, and error bodies carry documentation_url and suggested_action (14 of 15). Dated API version 2020-09-14 with a versioning page, and a changelog with nine dated entries since 2 July 2026 (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.

## 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: whether the Trial plan for new US and Canadian teams asks for a card
- The pricing page still advertises 200 free Production calls per product while the listing says the Trial plan replaced Limited Production; we didn't establish which applies to a team created today
- unchecked: GitHub issue and pull request responsiveness on plaid-node and plaid/mcp (the GitHub API wasn't available to us)

## Weaknesses

- No prices before a Production application, and Transactions, Liabilities and Investments bill per Item each month until removed
- Four plaid-node major versions between 23 July and 1 September 2026, each with breaking changes
- No Retry-After on 429 and no list of which error codes are safe to retry
- Six incidents marked major on the status page since 10 July 2026, five of them single banks
- One team secret reaches every product, Transfer included; there's no read-only key

## 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.

- Point calls at https://sandbox.plaid.com until you hold a Production secret; request shapes are the same
- Create a sandbox Item with /sandbox/public_token/create and ins_109508, then /item/public_token/exchange, no Link UI needed
- Use /transactions/sync with its cursor and back off on 429 RATE_LIMIT_EXCEEDED yourself, since no Retry-After comes back
- Call /item/remove when the user is done, or subscription products keep billing
- Send an idempotency_key on `/transfer/authorization/create` so a retried request can't authorise twice

## What the review panel asked for

- fewer sdk majors
- read-only secrets
- approval step for Transfer

## 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.