# Fix list: MagicBell From Anchor Terminal's listing at https://www.anchorterminal.com/tools/magicbell, the October 2026 research run, assessed 9 October 2026. Grade C, 58.8 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 MagicBell: 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. Security & auth, 50 out of 100, up to 8.8 more on the total Why it scored 50: Two credentials. A Project JWT is created in the dashboard with a name and an expiry, is shown once and can be discarded. A User JWT is signed by the customer's backend with HS256 and limits calls to one user. Project tokens carry no scopes (22). The User JWT confines a client to one user's inbox, but no read-only project token or confirmation for deletes was found, and `DELETE /users/{user_id}` removes a user and all associated data (6). Responses carry the project's own notifications plus event payloads from connected sources such as Stripe and GitHub, with no guidance on untrusted content (8). An events feed in the API and dashboard records deliveries with the HTTP request and response. No audit log of token or admin actions was found (8). The Trust Centre states GDPR compliance, AES-256 encryption and penetration testing. No SOC 2 or ISO 27001 statement, bug bounty or security.txt was found, and vulnerability reports go through a contact form (6). 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. ## 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). Plans and a unit price are public. Builder is $0 for 1,000 deliveries a month, Startup $249 for 50,000 and $0.0025 for each further delivery (20). The $0 plan needs only a sign-up, but the pricing page doesn't say whether a card is asked for and we did not create an account (15). A person signs up in a browser and creates the token in the dashboard, and the terms forbid accounts registered by automated methods (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, 63 out of 100, up to 6 more on the total Why it scored 63: List endpoints take `limit` (default 50) and there is no field selection (12). Cursor pagination with `starting_after` and `ending_before` on 21 endpoints, and notifications filter by status, category and topic (14). Errors return an array of `code` and `message` pairs, with no published list of codes (8). `Idempotency-Key` holds for 24 hours on some endpoints, which the docs don't name, and saves are PUT upserts (14). A broadcast needs only `title` and `recipients`, and official clients exist for JavaScript, Go, Swift and Java, with a CLI (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. Schema & documentation, 68 out of 100, up to 5.2 more on the total Why it scored 68: OpenAPI 3.1 file at /docs/api/openapi.json with 84 paths and 144 operations (25). /llms.txt returns 404 and no Markdown docs were found (0). Every operation has a description that states its purpose, many of them templated across the 16 integration types, and none says when not to use it (10). Required fields, patterns, length limits and ten enums are present and unknown query parameters are refused, but `custom_attributes` and workflow `input` are free-form objects and the file declares no security schemes (11). 88 schema examples, but only 200, 201 and 204 responses are described, and the docs show one error example with no list of codes (7). The path carries the version, v1 is marked deprecated and the changelog is public on GitHub Discussions (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. ## 5. Reliability, 75 out of 100, up to 5 more on the total Why it scored 75: Graded as a hosted API. Status page on Better Stack at status.magicbell.com with four components and daily history (20). In the 90 days to 9 October 2026 the API and dashboard show no downtime. In-app delivery shows about 10 minutes on 13 July, 11 minutes on 31 July and 39 minutes on 26 August, and email delivery 39 minutes on 26 August, with no incident write-up in the feed (20). The limit is published as 500 requests a minute per IP address (15). A 429 is documented with a 60-second block, and an `Idempotency-Key` header caches the first response for 24 hours on some endpoints, but no Retry-After header or backoff guidance was found (10). No SLA found on the pricing page or in the terms (0). API v2 is the current version (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. ## 6. Maintenance & community, 49 out of 100, up to 4.5 more on the total Why it scored 49: The last release is magicbell-js 1.8.0 on 29 July 2026, with a changelog entry the same day, 72 days before this check (20). One release or dated changelog entry since 11 July 2026 (0). A public changelog, GitHub Discussions and a Slack community exist. We did not test how fast questions are answered (8). Official clients are published for JavaScript (1.8.0, July 2026), Go (v0.5.0, March 2026), Swift and Java, and the CLI's last npm release is 1.4.0 of 8 January 2026 (13). The JavaScript repository runs test, lint and release workflows and took automated dependency updates up to 3 August 2026 (8). 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, 58 out of 100, up to 3.7 more on the total Made of editorial 30, provenance 86. Why it scored 58: Closed service under published terms. The Go client and the CLI are MIT, and the JavaScript client carries MagicBell's own Licence Agreement, which bars benchmarking the software (15). The privacy policy of 2 July 2024 covers visitors and account holders, gives no retention period beyond the time necessary, and does not address notification content or recipients. The Trust Centre says data is never sold or shared with third parties, while the policy marks account data as sold or shared with Google, X, Segment and Mixpanel. The Trust Centre mentions data processing agreements and none is published (6). API v1 and the old `@magicbell/cli` package are marked deprecated, with no policy or dates (5). The status page names Ably for in-app delivery and the policy names four analytics services with their countries. No sub-processor list or hosting region for the service was found (4). 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: 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 7 of the 8 things a reader expects (9.3 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: live error bodies and response headers, because api.magicbell.com/robots.txt disallows every path and no request was sent to the API - unchecked: whether sign-up for the Builder plan asks for a card, since we did not create an account - Which endpoints honour `Idempotency-Key`. The docs say some and name only notification creation as an example - Whether a 429 carries a Retry-After header. The docs state only a 60-second block - Whether a data processing agreement or sub-processor list exists. The Trust Centre mentions agreements and links none - unchecked: the body of the 29 July 2026 changelog entry, which the page did not show our reader, and whether community questions get answers - The status history is in two-day buckets, so the downtime dates are the start of each bucket ## Weaknesses - No batching or digest step found. Workflows have four commands (broadcast, wait, pause, abort) - No llms.txt or Markdown docs, and the OpenAPI file documents no error responses - Project tokens are project-wide with no scopes or read-only mode - No SLA, SOC 2 or ISO 27001 statement, security.txt or sub-processor list found - The Trust Centre says data is never sold or shared, while the privacy policy of 2 July 2024 marks account data as sold or shared ## 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. - Ask the owner for a Project JWT from the dashboard's Project Auth page. The terms forbid accounts registered by automated methods - Send `Idempotency-Key` with a UUID on `POST /v2/broadcasts`, and reuse it on a retry within 24 hours - Stay under 500 requests a minute per IP. A 429 blocks the address for 60 seconds, so put many recipients in one broadcast - Use a Project JWT for broadcasts, users, integrations and workflows, and a User JWT for one user's inbox, tokens and preferences - Read delivery failures from `GET /v2/events` and the broadcast's `status`, since email, SMS and push go through providers the owner connects ## 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.