Gumloop

# Fix list: Gumloop

From Anchor Terminal's listing at https://www.anchorterminal.com/tools/gumloop, the October 2026 research run, assessed 8 October 2026. Grade C, 61.6 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 Gumloop: 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, 20 out of 100, up to 10 more on the total

Why it scored 20: No x402, MPP or L402 (0). Per-unit prices are public without a login, $0.005 a credit, 5 credits a session-minute, 1 credit a connector call and an 8 per cent orchestration fee (20). No free plan is listed and the 14-day trial needs a card (0). A person signs up in a browser, adds a card and creates the key in the app (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, 50 out of 100, up to 8.8 more on the total

Why it scored 50: Personal and team API keys, and OAuth 2.0 with PKCE, refresh tokens and a revoke endpoint, where the one API scope, `gumloop_api`, grants the whole developer API. Key rotation was not found in the pages read (20). Less 10 because the docs name the `api_key` query parameter the default method, and a webhook trigger's secret sits in its URL (10). Approval modes per connector and per tool, CEL app rules and an approvals endpoint, with Always allow as the default and no read-only API key (15). Agents read third-party content through connectors and a browser, and no prompt-injection guidance was found in the pages read (3). Audit logs through `GET /get_audit_logs` and connector activity logs, with retention set by the Enterprise agreement (12). The security page claims SOC 2 Type II and HIPAA. No security.txt, no bug bounty found, and the trust centre was not read (10).

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. Agent ergonomics, 66 out of 100, up to 5.5 more on the total

Why it scored 66: List operations take `page_size`, and a session read returns its whole message list. The MCP server has 46 tools, with no toolsets found (15). Cursor pagination on 13 operations with search, team, status and sort filters (18). Named error codes such as `gummie_rate_limit` and `trigger_type_not_editable`, and MCP calls return a per-result `code`, `message` and `type`, but the API has no single error format (13). A caller-supplied `session_id`, idempotent attach and detach calls and an agent `version` check on update. No idempotency key, and no MCP tool annotations found in the docs (12). The Python SDK and CLI are current. The npm package dates from 30 December 2024 and runs flows only (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. Transparency & trust, 57 out of 100, up to 3.8 more on the total

Made of editorial 41, provenance 72.

Why it scored 57: Closed service under published terms with AgentHub Inc., and the SDK is Apache-2.0 (15). The privacy policy and terms agree that self-serve content may be used to train Gumloop's models and Enterprise content is not. Retention is 'as long as is necessary' with no periods, and DPAs are for Enterprise on request (14). Flows are labelled Legacy on the pricing page and the SDK's flows client raises a deprecation warning, with no dated policy or end date found (6). The privacy policy says data is processed in the United States. No subprocessor list was read, because the trust centre is drawn by script (6).

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: gumloop.com, registered 2024-04-25 (2 years) (7 of 15)
- Terms of service: read, states 5 of the 7 things a reader expects, and has 3 clauses that cost points (2.3 of 10)
- Privacy policy: read, states 7 of the 8 things a reader expects, and has 1 clause that costs points (7.3 of 10)
- security.txt: not found (0 of 10)

## 5. Reliability, 82 out of 100, up to 3.6 more on the total

Why it scored 82: Graded as a hosted service on the REST API. Status page on incident.io with Platform and API components and a history view (20). The history shows no incident reported for July, August, September or October 2026 (30). Limits are published as concurrency, 25 agent interactions on Pro and 100 on Enterprise, and 100 requests a minute per webhook trigger. No per-request limit for the API was found (12). On Pro a request over the limit gets HTTP 429 with `gummie_rate_limit`, webhook 429s carry `Retry-After`, and a caller-supplied `session_id` returns 409 on a repeat. No backoff guidance for the API was found (10). No SLA found on the pricing page, the terms or the docs (0). The API is not marked beta or preview (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. Schema & documentation, 83 out of 100, up to 2.8 more on the total

Why it scored 83: Public OpenAPI 3.0.0 file with 105 operations on 83 paths (25). `llms.txt`, `llms-full.txt` and a Markdown copy of every docs page (10). 98 of 105 operations carry a description, and many say what is idempotent, what merges and what replaces (16). Typed parameters with 80 enums, but only 27 named schemas and free-form `metadata` objects on agents (10). cURL and Python samples and status codes on most operations, with no shared error schema (11). The path is versioned at `/api/v1` and the product changelog is public and dated, but the file's version stays at 1.0.0 and there is no changelog for the API alone (11).

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, 77 out of 100, up to 2 more on the total

Why it scored 77: Release 12.1.0 is dated 7 October 2026 and the Python SDK 0.5.5 was published on 8 October (30). 28 dated changelog releases between 9 July and 7 October 2026 (20). Public changelog, a forum, a help centre and support by email and ticket form. We did not test a support channel (10). The Python SDK is current but the npm package has had no release since 30 December 2024, and no entry was found in the official MCP registry (10). The SDK repository runs ruff, pyright and pytest in CI and locks dependencies. It has no changelog file (7).

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.

- 2026-10-08. The API getting-started page says the Python and JavaScript SDKs 'wrap all of it'. The JavaScript SDK page documents only `runFlow`, and the npm package `gumloop` is 1.0.2 from 30 December 2024 (https://docs.gumloop.com/api-reference/getting-started.md, https://registry.npmjs.org/gumloop). 2 points for a misleading claim.

## 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: the trust centre at trust.gumloop.com is drawn by script, so certifications, the subprocessor list and any disclosure contact there were not read
- unchecked: the MCP server's tool definitions and annotations, which are only visible after signing in. The count of 46 is the docs' figure
- unchecked: GitHub stars, open issues and how quickly issues are answered on gumloop/gumloop-py
- unchecked: the pricing page's FAQ answers, which the page did not show our reader
- The terms of service forbid access by any robot, spider or other automatic device. We read the public site pages that robots.txt allows before reaching that clause, and read nothing further on www.gumloop.com after it
- The terms give the date as 11/06/2026, which we read as 11 June 2026
- Whether API keys can be rotated or limited in scope was not found in the pages read
- No free plan appears on the pricing page, though the privacy policy mentions free accounts

## Weaknesses

- API keys and the `gumloop_api` OAuth scope need the Pro plan at $37 a month. The 14-day trial needs a card
- The docs call the API key in the URL query string the default method, and a webhook trigger's secret is part of its URL
- The privacy policy of 11 June 2026 says self-serve content, Pro and trials included, may be used to train Gumloop's AI models
- The npm package `gumloop` is still 1.0.2 from 30 December 2024 and only runs legacy flows, while the docs say both SDKs wrap the whole API
- No SLA and no per-request rate limit found. On Pro, a 26th concurrent run is rejected with HTTP 429 and skipped for triggers

## 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 the key as `Authorization: Bearer` and, for a personal key, add the `x-auth-key` header with the user ID. Do not put `api_key` in the URL
- Create sessions with `POST /agents/{agent_id}/sessions`, then poll `GET /sessions/{session_id}`. A 202 means processing or queued, and the reply is not in the response
- Supply your own `session_id` when creating a session. A repeat returns 409, which makes a retry safe
- For streamed output send the same body with `stream: true` to `ws.gumloop.com`. `api.gumloop.com` answers 400 to it
- Treat a webhook trigger URL as a password. It returns `{"success": true}` at once and never carries the agent's output

## 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.