Vapi API + MCP by Vapi

# Fix list: Vapi API + MCP

From Anchor Terminal's listing at https://www.anchorterminal.com/tools/vapi, the October 2026 research run, assessed 1 October 2026. Grade B, 63.7 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 Vapi API + MCP: 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, 35 out of 100, up to 8.1 more on the total

Why it scored 35: No x402, MPP or L402 (0 of 40). The $0.05 hosting fee and Twilio transport rates are public and provider costs pass through at cost, but the total per minute depends on choices the pricing page only estimates (15 of 20). $5 of credit and no card needed to start, per last week's check (20). Access starts with a human signup (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. Security & auth, 58 out of 100, up to 7.4 more on the total

Why it scored 58: Private keys for the REST API and MCP server, and public keys that can be limited to allowed origins and allowed assistants. We found no rotation or revocation guidance (25). The docs advise one key per environment, but there's no read-only private key and the MCP tools that spend money carry no annotations (8 of 20). Assistants listen to callers and we found no prompt-injection guidance (5 of 15). Call logs carry a cost breakdown per provider and, since 21 September 2026, provider request IDs for BYOK calls, but we found no audit log of account actions (8 of 15). HIPAA, GDPR and PCI pages and a SOC 2 Type II claim in the FAQ. No security.txt or bug bounty found. The June 2026 npm incident was written up in public (12 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.

## 3. Reliability, 70 out of 100, up to 6 more on the total

Why it scored 70: Statuspage at status.vapi.ai with component history (20). Thirteen incidents since 3 July 2026, most of them under 40 minutes or planned maintenance. On 12 August call failures ran for 2 hours 3 minutes, and a second call-failure incident on 19 August has no published duration, so we count one major (10). Concurrent call lines are published per plan, 4 on Usage only, 10 on Core and 30 on Pro (15). When lines are full the call queues and the response's `subscriptionLimits` sets `concurrencyBlocked`, which an agent can read, but we found no request rate limits, Retry-After or idempotency guidance (5 of 15). The Pro package carries a 99 per cent uptime SLA and Premier 99.9 per cent (10). The 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. Agent ergonomics, 64 out of 100, up to 5.9 more on the total

Why it scored 64: The MCP server README lists 20 tools for assistants, calls, phone numbers, tools and login (15 of 25). Call and assistant lists take limits and date filters (17 of 20). Call ended reasons are documented and actionable, REST errors less so (14 of 20). We found no idempotency keys, and the README shows no `readOnlyHint` or `destructiveHint` annotations, though `vapi_create_call` places real calls and `vapi_buy_phone_number` spends money (3 of 20). Server SDKs for TypeScript and Python plus web and mobile clients (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, 76 out of 100, up to 2.1 more on the total

Made of editorial 70, provenance 82.

Why it scored 76: Closed platform with clear terms, and the MCP server is MIT (18 of 30). Raw data retention is published per plan, 14 days on Usage only, 30 on Core and 180 on Pro, with a zero data retention option and HIPAA, GDPR and PCI pages that agree (22 of 30). Workflows were retired on 18 August 2026 with a migration guide, and model retirements are emailed since 7 September (18 of 20). A data-flow page names where call data goes, but we didn't find a subprocessor list (12 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: vapi.ai, registered 2023-08-22 (3 years) (7 of 15)
- security.txt: not found (0 of 10)

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

Why it scored 89: OpenAPI at api.vapi.ai/api-json (25). llms.txt with Markdown copies of more than 500 pages (10). Guides explain when to use squads, tools and fallback plans (15 of 20). The OpenAPI types every field, though assistant configuration is a large nested object (12 of 15). Examples throughout and a call-errors guide that explains ended reasons, with less on REST error bodies (12 of 15). A weekly changelog and, since 17 August 2026, versioned assistants and tools with draft and publish (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.

## 7. Maintenance & community, 88 out of 100, up to 1.1 more on the total

Why it scored 88: Changelog entry on 21 September 2026 (30). Weekly entries, eleven since 13 July (20). The MCP server repository has 4 open issues and 11 open pull requests, and we didn't check reply times (15 of 25). Server SDKs, web SDK and MCP server on npm and PyPI (15). The MCP server is MIT with GitHub Actions (8 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.

## Deductions

Each comes off the total. A fixed and documented problem counts for less at the next check.

- 2026-06-03. A compromised developer GitHub token was used to push malicious code to Vapi repositories and publish four malicious `@vapi-ai/server-sdk` versions (0.11.1, 0.11.2, 1.2.1, 1.2.2) to npm. Vapi says they were removed within about three hours with zero downloads and no customer data exposed, and it published a write-up. Fixed and documented, so we deduct 4 of a possible 15 (https://vapi.ai/blog/our-response-to-june-3-supply-chain-incident).

## 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.

- The listing's toolCount of 10 is stale. The MCP server README lists 20 tools, and we didn't check whether the hosted server exposes the same set.
- Duration of the 19 August 2026 call-failure incident.
- REST request rate limits and whether 429 responses carry Retry-After.
- The $5 no-card start comes from last week's check, not this run.

## Weaknesses

- Real per-minute cost depends on provider choices
- Usage only has 4 concurrent lines and 14 days of retention
- MCP tools that place calls or buy numbers carry no annotations
- No request rate limits or Retry-After guidance found
- Malicious SDK versions reached npm for about three hours on 3 June 2026

## 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.

- Read `subscriptionLimits.concurrencyBlocked` in the `POST /call` response, a full account queues the call rather than failing
- Confirm with a person before `vapi_create_call` or `vapi_buy_phone_number`, the server won't ask
- Pin transcriber, model and voice and set fallback plans so one provider outage doesn't drop calls
- Read the call log's cost breakdown to see what each provider charged
- Check `@vapi-ai/server-sdk` isn't pinned to 0.11.1, 0.11.2, 1.2.1 or 1.2.2

## What the review panel asked for

- publish REST rate limits
- say whether 429 carries Retry-After
- read-only private keys
- destructive hints on MCP tools

## 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.