Ramp

# Fix list: Ramp

From Anchor Terminal's listing at https://www.anchorterminal.com/tools/ramp, the October 2026 research run, assessed 8 October 2026. Grade C, 57.3 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 Ramp: 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. Reliability, 30 out of 100, up to 14 more on the total

Why it scored 30: Read with the hosted lines and scored on the Developer API, the surface an outside agent calls with a token. No public status page was found. status.ramp.com doesn't resolve, and no status link appears in the docs export, the help centre article, the trust centre or the home and pricing pages we read (0). With no page there is no incident record to read (0). The limit is published, 200 requests per rolling 10 seconds per source IP, with a 60-second request timeout (15). The docs ask for exponential backoff of 1, 2 and 4 seconds on 429 and on 504. `X-Idempotency-Key` is accepted on 8 of 155 write operations (funds, procurement requests, vault cards, accounting syncs), and a Retry-After header isn't documented (9 of 15). No SLA was found in the Platform Agreement or the API Agreement (0). The API is v1, but 89 of 260 operations carry `x-beta` and the MCP guide says the server is subject to change (6 of 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.

## 2. Payments & pricing, 25 out of 100, up to 9.4 more on the total

Why it scored 25: Read with the hosted rubric. Ramp's API isn't paid for by x402, MPP or L402. Ramp does let an agent pay other sellers by x402 from a funded wallet or by Stripe MPP, which is a product capability and not a way to buy access (0). Plan prices are public, Free at $0 per user per month and Plus at $15 per user per month plus a platform fee by team size that isn't stated, with Enterprise on request. The API has no per-call price (10). The Free plan needs no payment card, but it needs an approved Ramp business account, and sandbox access is by request form, so partial credit (15 of 20). The Developer MCP server answers docs questions with no account. Account data needs an admin to create an app in the dashboard or a person to complete an OAuth sign-in (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. Agent ergonomics, 66 out of 100, up to 5.5 more on the total

Why it scored 66: `page_size` sizes a page (default 20, maximum 100) and `functional_currency_amount` is opt-in on transactions. No field selection or expand parameter was found in the spec (18 of 25). Keyset pagination returns the next page as a full URL in `next`, and 50 list operations take `page_size`, with filters by department, card, merchant, state and date on transactions. The docs say bills and transactions have no updated-at filter (16 of 20). Errors arrive as `error_v2` with `error_code`, `message`, `additional_info` and `error_id`, and every response carries `x-trace-id`. There is no published list of codes, and the docs say unknown fields are ignored on some endpoints and rejected on others (14 of 20). `X-Idempotency-Key` on 8 of 155 write operations, and asynchronous user creation returns a task ID to poll. The spec marks 14 operations with `x-read-only` and 3 with `x-destructive`. The MCP server's own annotations couldn't be read without an account (10 of 20). List calls need no required parameters. No official SDK was found in the docs. The MIT ramp-cli prints JSON in agent mode and has a dry-run flag (8 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.

## 4. Security & auth, 79 out of 100, up to 3.7 more on the total

Why it scored 79: OAuth 2.0 with client credentials or the authorisation code grant, about 80 scopes in the spec in `resource:read` and `resource:write` form, tokens bound to scopes at issue and sent in the `Authorization` header, a revoke endpoint, and PKCE (S256) on the MCP server (30). Read scopes are separate from write scopes, card numbers need `cards:read_vault` and a production review by Ramp, and MCP and CLI users act with their own Ramp permissions. Admins on Plus choose who may connect. A client credentials token acts with the app creator's permissions, MCP tools can approve transactions and reimbursements for an approver, and no confirmation step was found for API writes (15 of 20). Memos, comments, merchant names and receipts are untrusted text, and no prompt-injection guidance was found (3 of 15). `GET /developer/v1/audit-logs/events` filters by user, date, event type and actor type (Plus plan), responses carry a trace ID, and MCP calls are attributed to the signed-in user (13 of 15). A valid security.txt expiring 18 March 2027, SOC 1 and SOC 2 Type 2 reports, ISO 27001:2022, ISO 42001, PCI DSS and a responsible disclosure entry on the trust centre. No bounty is stated (18 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. Transparency & trust, 70 out of 100, up to 2.6 more on the total

Made of editorial 55, provenance 84.

Why it scored 70: Closed service with terms at stable URLs, the Platform Agreement (updated 22 September 2026) and the API Agreement (26 March 2026). The CLI is MIT (16 of 30). The privacy policy (21 August 2026) describes retention by criteria without fixed periods and says Ramp will give notice and obtain permission before training a general-purpose model on company content. The DPA (10 July 2026) deletes personal data on written request after termination, with back-ups kept to Ramp's own schedule (20 of 30). The changelog labels deprecated fields, but the API Agreement says Ramp may change or discontinue the APIs with or without notice, and no notice period is published (6 of 20). The trust centre lists sub-processors and key service providers and posts additions with name, location and purpose, such as Modal in the United States. The full documents sit behind an access request and NDA (13 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):

- Terms of service: published, but our reader couldn't read it (7 of 10)
- Privacy policy: published, but our reader couldn't read it (7 of 10)
- Status page: not found (0 of 10)

## 6. Schema & documentation, 86 out of 100, up to 2.3 more on the total

Why it scored 86: A public OpenAPI 3.0.2 spec with 260 operations at docs.ramp.com/openapi/developer-api.json (25). llms.txt, a plain-text file per guide, llms-api.txt and llms-full.txt (10). Every operation has a summary or description, and the 133 agent-tool operations in the CLI's bundled spec open with when to use the tool. Developer API descriptions are shorter, and few say when not to call an endpoint (14 of 20). 638 enums, many with `x-enum-descriptions`, UUID and date-time formats, and `x-relationship` links between resources. Error `additional_info` is a free-form object (13 of 15). 264 examples. 400 is documented on 75 operations, 404 on 17, 401 on 3 and 429 on none, and the error guide lists eight status codes without a catalogue of `error_code` values (9 of 15). The path is versioned v1, and the changelog is dated to the day with an RSS feed (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, 74 out of 100, up to 2.3 more on the total

Why it scored 74: The changelog's latest entry is 7 October 2026, and ramp-cli v0.2.54 was tagged on 6 October (30). 26 dated changelog entries between 15 July and 7 October 2026, and 48 CLI tags since 10 July (20). Closed service with a daily-dated changelog and RSS feed. Support is a ticket raised from the developer console and answered by email, and the CLI's public repository shows 7 open issues and pull requests. Response times couldn't be observed (10 of 15). No official SDK was found, and Ramp's MCP server isn't in the official MCP registry, where the only Ramp entry is a third party's (com.mcparmory/ramp). The official CLI is current (6 of 15). The CLI repository runs lint, pytest and a smoke test in CI and pins dependencies in uv.lock. It installs from a script or Homebrew, not PyPI (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.

- 2026-06-25 and 2026-02-11. The changelog records the webhook event type `transactions.all_requirements_met_and_approved` removed on 25 June 2026, three days after the 22 June entry marked it a deprecated alias, and `functional_currency_amount` dropped from default transaction responses on 11 February 2026. Both are documented on the day they shipped and the first names a replacement. Notice by email couldn't be checked, so the smallest deduction applies (https://docs.ramp.com/developer-api/v1/changelog).

## 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 Ramp publishes a status page at an address other than status.ramp.com, which doesn't resolve. None was linked from the pages we read
- unchecked: the live Ramp MCP server's tool list, input schemas and readOnlyHint and destructiveHint annotations, which need a signed-in Ramp account. The count of 120 comes from the CLI's bundled spec
- unchecked: whether 429 responses carry a Retry-After header
- unchecked: whether customers were told by email before the 25 June 2026 webhook event removal and the 11 February 2026 response change
- unchecked: the full sub-processor list and the security reports on trust.ramp.com, which need an access request and NDA
- unchecked: whether an SLA exists in Enterprise order forms. None was found in the Platform Agreement or the API Agreement
- unchecked: whether the Free plan includes Developer API access. The spec marks 209 operations as not needing Plus, and the pricing page doesn't list the API under any plan
- The Claude Code and JSON config lines in the MCP guide are stripped from the plain-text export, so `connect` holds only the install script and the documented first request
- The x402 guide linked from the cards guide wasn't read. Ramp's x402 support is for paying sellers from a funded wallet, not for paying for Ramp's API
- The trust centre's phishing advisory against Ramp customers is undated in the text we read and concerns a third party's campaign, so no deduction was made

## Weaknesses

- No public status page was found. status.ramp.com doesn't resolve, and no status link appears in the docs, help centre or trust centre pages read
- 89 of 260 operations carry `x-beta`, among them creating a fund and splitting a transaction, and the MCP guide says the server is subject to change
- `X-Idempotency-Key` is accepted on 8 of 155 write operations. Creating a bill, a purchase order or a vendor has none
- The API Agreement says Ramp may change or discontinue the APIs at any time with or without notice
- No official SDK was found in the docs, and 51 operations (custom records, purchase orders, the audit log) need the paid Plus plan

## 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.

- Send `scope` on the client credentials token request. Without it Ramp issues a token with no scopes and every resource call fails with 403
- Follow the full URL in `next` to page. `page_size` defaults to 20 and stops at 100, and the cursor is opaque
- Stay under 200 requests in any 10 seconds per source IP and back off 1, 2 then 4 seconds on 429. No Retry-After header is documented
- Don't use `synced_after` or `from_created_at` to find changed records. Bills and transactions have no updated-at filter, so subscribe to webhooks
- Send `X-Idempotency-Key` when creating funds or procurement requests. Elsewhere a retried POST can create a duplicate, so read back before retrying

## 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.