Upstash QStash

# Fix list: Upstash QStash

From Anchor Terminal's listing at https://www.anchorterminal.com/tools/upstash-qstash, the October 2026 research run, assessed 8 October 2026. Grade BB, 72.3 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 Upstash QStash: 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, 40 out of 100, up to 7.5 more on the total

Why it scored 40: No x402, MPP or L402 (0). $1 per 100,000 messages and $0.05 per GB over 50 GB, published without login (20). Free plan with 1,000 messages a day and no card required per the pricing page (20). A person signs up in the browser and copies the token from the console. No keyless or programmatic route to a first QStash token was found (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, 61 out of 100, up to 6.8 more on the total

Why it scored 61: One full-access token and one read-only token per region, revoked by a reset, with two signing keys that rotate by API. No scopes per queue, destination or schedule (24 of 30). The token is accepted in a `qstash_token` query parameter as a documented option (less 10). Read-only token, read-only Developer API keys and MCP grants, and destructive labels on MCP tools, though the read-only token still reads message bodies (18 of 20). Logs and the dead letter queue return message bodies and endpoint responses. Field redaction is documented, and no prompt-injection guidance was found (5 of 15). Per-message logs by API, kept 3 to 14 days, and Prometheus and Datadog export on Prod Pack. No account audit log was found (10 of 15). Valid security.txt, a disclosure policy with safe harbour and rewards for critical findings, and SOC 2 Type 2 for QStash only with Prod Pack. ISO 27001 is in progress (14 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. Schema & documentation, 78 out of 100, up to 3.6 more on the total

Why it scored 78: Public OpenAPI 3.1 file with 43 operations (25). llms.txt and a Markdown copy of every docs page (10). Each header parameter says what it does with formats and examples, and the docs point from queue parallelism to flow control, but few entries say when not to use an endpoint (14 of 20). Enums on method and message state, typed query parameters on logs, but most options are string headers, among them the flow-control value and retry-delay expression (10 of 15). curl, TypeScript and Python examples on each page and status codes per operation, with one error schema holding a single string and no error list (10 of 15). The path is versioned at /v2, but the docs changelog stops at February 2026 and the GitHub Discussions list that replaced it shows nothing released after March 2026 (9 of 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.

## 4. Reliability, 83 out of 100, up to 3.4 more on the total

Why it scored 83: Graded on the hosted REST API. status.upstash.com is a Statuspage with a QStash component split into EU-CENTRAL-1 and US-EAST-1 and a full incident history (20). Two QStash incidents since 10 July 2026, both in us-east-1 and both short. Publish errors on 16 July were fixed within minutes, and a message persistence fault on 28 August ran from 22:20 to 22:33 UTC and was marked major (20 of 30). Publish, enqueue and batch have no per-second limit, daily message caps and parallelism are published per plan, but the burst limit on management endpoints has no published number (10 of 15). 429 responses carry `Burst-RateLimit-Reset`, the SDK raises typed rate-limit errors with a reset time, and deduplication IDs make a repeated publish safe (13 of 15). 99.99% SLA with the Prod Pack add-on or Enterprise (10). 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.

## 5. Agent ergonomics, 83 out of 100, up to 2.8 more on the total

Why it scored 83: Graded as an API. Publish returns only a message ID, logs take `count` up to 100, and there's no field selection. Through MCP, `?features=qstash_workflow` cuts 55 tools to 10 (17 of 25). Cursor pagination and filters for state, URL, queue, schedule, label and date on logs and the dead letter queue (20). Errors are an HTTP status plus one message string, such as `unable to authenticate: token is empty`, with typed SDK errors for rate limits and no code list (12 of 20). Deduplication IDs with a 10-minute window, 202 for a duplicate, `Upstash-Message-Id` for idempotent receivers, and read, write or destructive labels on MCP tools (20). One curl with a token publishes a message, with 3 retries by default. Official SDKs for TypeScript and Python, the Python one last released in March 2026 (14 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.

## 6. Maintenance & community, 77 out of 100, up to 2 more on the total

Why it scored 77: `@upstash/qstash` 2.12.0 reached npm on 29 September 2026 (30). Three releases in 90 days, 2.11.2 on 14 July, 2.11.3 on 22 July and 2.12.0 (20). Public changelog, Discord and email support, with the docs changelog last updated for February 2026 and GitHub issue replies not read (8 of 15). The TypeScript SDK is current, the Python SDK's last release is 3.4.0 on 18 March 2026, and the MCP server is in the official registry as io.github.upstash/mcp-server (11 of 15). CI and test workflows on both SDK repositories, and the TypeScript SDK replaced crypto-js with Web Crypto in September 2026 (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.

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

Made of editorial 62, provenance 100.

Why it scored 81: Closed service with published terms. The SDKs and MCP server are MIT, and the local development server has its own testing-only licence (18 of 30). DPA, privacy policy, subprocessor list and security measures are public, and log and dead letter retention is published per plan. The security measures PDF says SOC 2 Type II is in progress while the QStash page says it's held with Prod Pack, and the Markdown pricing page gives two answers on retry billing (22 of 30). No deprecation policy found. Queue parallelism is to be deprecated 'at some point' with no date (5 of 20). Subprocessors are listed with addresses, the two regions are named, and the stdio MCP server's telemetry is documented with `--disable-telemetry` (17 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.

## 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.

- Whether retries are billed. The pricing FAQ says each delivery attempt is a message, and the Markdown version of the same page also says retries are free
- unchecked: the price of the Prod Pack add-on, which wasn't on the pricing page we read
- unchecked: the numeric burst limit on management endpoints such as logs and the dead letter queue
- unchecked: the hosted MCP server's QStash tool schemas, which aren't published. The stdio package's source is public and has a different tool list
- unchecked: issue response times on the SDK repositories, because the GitHub API refused our requests and issues aren't in a clone
- Whether the February 2026 change to the GET /v2/globalParallelism response was announced beforehand
- Whether SOC 2 Type 2 is complete for QStash. The QStash page says yes with Prod Pack, the security measures PDF says in progress, and the compliance page names only Redis
- The first release date is the npm package's creation date, 14 June 2022, not a vendor statement about the service

## Weaknesses

- One full-access token and one read-only token per region. No per-queue or per-destination scopes were found
- The token is accepted as a `qstash_token` query parameter, which the webhook receiver guide relies on
- The Markdown pricing page says retries are free and, in its FAQ, that each retry is billed as a message
- Two QStash incidents in us-east-1 in 90 days, on 16 July and 28 August 2026, both under 15 minutes
- The docs changelog stops at February 2026, and the Python SDK last shipped on 18 March 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.

- Use the regional host that matches the token. `qstash.upstash.io` is the EU region, and US tokens work only on `qstash-us-east-1.upstash.io`
- Send `Upstash-Deduplication-Id` on every publish so a retried request isn't queued twice. The window is 10 minutes
- Budget for retries. Per the pricing FAQ each delivery attempt is billed as a message, so set `Upstash-Retries` deliberately
- Give monitoring agents the read-only token, and set `Upstash-Redact-Fields` on publish, because that token still reads message bodies and headers
- Make the destination idempotent on `Upstash-Message-Id`. Delivery is at least once, and duplicates can follow a server restart

## 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.