Devin

# Fix list: Devin

From Anchor Terminal's listing at https://www.anchorterminal.com/tools/devin, the October 2026 research run, assessed 8 October 2026. Grade C, 55.7 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 Devin: 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: Hosted reading. Status page at www.devinstatus.com with 14 components and incident history (20). In the 90 days to 8 October 2026 it lists three critical incidents (22 July, 43 minutes of platform-wide queuing, 13 August, Devin Cloud unavailable for 2 hours 40 minutes, 24 September, the web app down for 2 hours) and six major ones on session start-up, so several majors (0). 429 is a documented response, but no rate limit figures were found (0). No Retry-After or backoff guidance found, and v3 session creation has no idempotency key, although the legacy v1 API had one (0). The platform terms supply the service as is with no service level, and enterprise terms weren't read (0). The v3 API is the current, recommended surface, with 44 of 239 operations marked beta under /v3beta1 (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, 20 out of 100, up to 10 more on the total

Why it scored 20: Published rubric. No payment protocol (0). Plan prices are public, $20, $200 and $40 a seat with an $80 team minimum, but usage past the quota is billed at an unpublished rate (10). A Free plan exists. The pricing page lists cloud agents from Pro upward, and we couldn't confirm API use on Free or whether a card is needed (10). A person signs up in a browser and provisions the service user in settings (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, 62 out of 100, up to 6.2 more on the total

Why it scored 62: API and MCP reading. The MCP server documents 13 tools, several of them consolidated manage tools, and list endpoints take `first` to size a page (17). Cursor pagination on every v3 list endpoint and filters on session search by tag, playbook, origin, user and time (18). problem+json with `error_code`, field-level errors on 422 and a 400 that lists valid platform names, but no published code list (15). No idempotency key on v3 session creation. `max_acu_limit` caps what a session can spend, and the MCP tool annotations couldn't be read without a key (4). Only `prompt` is required to start a session. No official SDK found, only a Terraform provider (8).

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, 72 out of 100, up to 4.9 more on the total

Why it scored 72: Hosted reading. Service-user keys with roles and named permissions per endpoint, revocable, with rotation endpoints in beta for Enterprise and expiring personal tokens. Custom roles are an Enterprise option, and Teams has Admin and Member only (25). Security profiles restrict network, MCP servers, git and the Devin MCP to read-only, and read-only API permissions exist. They are opt-in, and session creation accepts `bypass_approval` (16). AI Guardrails screen user messages and pull request comments for prompt injection, on Enterprise only, and they don't cover web pages or repository content Devin reads (8). Audit logs through the API need an enterprise-level permission. Session events and insights are available on all plans (10). SOC 2 Type II and ISO/IEC 27001:2022, a disclosure address at security@cognition.ai, no security.txt, and the trust centre answers No to having a bug bounty (13).

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. Schema & documentation, 80 out of 100, up to 3.3 more on the total

Why it scored 80: API reading. Public OpenAPI 3.1 spec for v3 with 239 operations, and separate v1 and v2 specs (25). llms.txt with 376 lines and a Markdown twin of every page (10). Each endpoint page names its required permission, but 61 of 239 operations have no description and the MCP tool descriptions are one line each (10). Enums on fields such as `devin_mode`, `origin` and `category`, with most strings unconstrained and nullable (9). curl examples in the quick starts, and a problem+json error schema declared on every operation, without a catalogue of `error_code` values (11). Versioned paths (v1, v2, v3, v3beta1) and a dated API release notes page (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.

## 6. Maintenance & community, 63 out of 100, up to 3.2 more on the total

Why it scored 63: Closed service. Newest release note dated 7 October 2026 (30). Nineteen dated release notes between 7 August and 7 October 2026, and API notes for August and September (20). Public release notes and a support address, with no public tracker or forum found (10). No official SDK found and the Devin MCP server didn't appear in a search of the official MCP registry (0). No public CI or packages to assess, with the docs and three OpenAPI specs kept current (3).

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, 69 out of 100, up to 2.7 more on the total

Made of editorial 61, provenance 77.

Why it scored 69: Editorial half. Closed source with clear terms from Cognition AI, Inc. (15). Terms, privacy policy, security page and data processing agreement agree with each other. Customer data may be used for model training by default on self-serve plans, with an opt-out on paid plans that also turns on zero data retention at model providers. Retention is stated as the length of the customer relationship, with no periods (18). Dated notices for the schedules endpoints (announced 10 September, creation refused from 24 September 2026) and the v2 clone endpoint, but no deprecation policy and no end date for v1 and v2 (10). Sub-processors listed with purpose and country, and 15 days' notice of additions (18).

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: devin.ai, registered 2022-12-06 (3 years) (7 of 15)
- Terms of service: read, states 6 of the 7 things a reader expects, and has 1 clause that costs points (7.1 of 10)
- Privacy policy: read, states 8 of the 8 things a reader expects, and has 1 clause that costs points (8 of 10)
- 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: security incidents or advisories in the last 12 months. Web search wasn't available, and no advisory page was found on the vendor's sites, so the deduction is 0 on what we could read.
- unchecked: whether the Free plan can create sessions through the API, and whether signup asks for a card. The pricing page and the docs describe Free differently.
- unchecked: the MCP server's tool schemas and annotations, which need a key. mcp.devin.ai/mcp answered 406 to a plain GET.
- unchecked: Enterprise Terms of Service, including any service level, and the documents on trust.cognition.ai, which need an NDA.
- No rate limit figures, on-demand unit rate or end date for the v1 and v2 APIs were found in the reviewed documentation.
- Devin CLI and Devin Desktop (formerly Windsurf) are separate local products on the same plans and weren't graded here.

## Weaknesses

- www.devinstatus.com lists three critical incidents (22 July, 13 August, 24 September 2026) and six major ones on the cloud agent or web app since 10 July 2026
- No rate limit figures, Retry-After or backoff guidance found in the API docs, and v3 session creation has no idempotency key
- Customer data may be used for model training by default on self-serve plans. The opt-out is for paid plans only, per section 3.3.1 of the platform terms
- A person must sign up and provision the service user in the web app. No key-creation route exists for an agent starting from nothing
- No official SDK found, and the Devin MCP server was not found in the official MCP registry

## 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.

- Use a `cog_` service-user key with the Member role. Legacy `apk_` keys fail against v3 and the MCP server with 401 or 403
- Set `max_acu_limit` on every session you create. Usage is metered by the work done and has no published unit rate
- Session creation is not idempotent in v3. After a timeout, list sessions by tag before creating again
- Enterprise keys and personal access tokens must send `X-Org-Id` to the MCP server. Organisation-scoped keys resolve it automatically
- Create scheduled work as an automation. POST to the schedules endpoint returns 403 for migrated organisations since 24 September 2026

## 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.