Scrapfly by Scrapfly (Joam Intelligence, LLC)

# Fix list: Scrapfly

From Anchor Terminal's listing at https://www.anchorterminal.com/tools/scrapfly, the October 2026 research run, assessed 1 October 2026. Grade B, 69.8 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 Scrapfly: 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, 50 out of 100, up to 10 more on the total

Why it scored 50: scrapfly.io/status reports an overall state (All Systems Operational on 1 October) with no components and no history, and status.scrapfly.io, which llms.txt links, still doesn't resolve in DNS (10 of 20). No readable incident history (5). Concurrency per plan is published (5 on Discovery to 100 on Enterprise) and each project has a configurable per-minute throttle (15). The error catalogue marks every code retryable or not and has three ERR::THROTTLE codes, but we found no Retry-After header or backoff guidance (10 of 15). The terms disclaim uptime and no SLA is published (0). The Scrape API is GA. Vault and alerting are beta, outside the main surface (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, 40 out of 100, up to 7.5 more on the total

Why it scored 40: No x402, MPP or L402 (0). Credit costs per request type (1 datacentre, 5 with rendering or Unblocker, 25 residential, 60 full-page screenshot) and plan prices are public. Unblocker surcharges per target aren't published, but cost_budget caps them (20). 1,000 free credits, no card, no time limit (20). A person signs up in a browser. OAuth still needs an account (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. Security & auth, 70 out of 100, up to 5.3 more on the total

Why it scored 70: OAuth 2 on the hosted MCP with dashboard revocation, open to all plans. API keys are per project, live or test, and since 15 April 2026 can be restricted to products (30). The canonical API form puts the key in the key query parameter and the MCP also accepts ?key=, so less 10 (20 of 30). Product-restricted keys, confirm=true on every mutating vault and alert tool, info_api_key off by default, and the self-hosted HTTP server binds loopback by default (15 of 20). The source gates mutating vault tools behind a preview step because prompt injection can reach them and redacts keys from CDP URLs. We found no user-facing guidance on untrusted page content (8 of 15). Monitoring dashboards, per-domain metrics and alerting. We didn't confirm per-request logs today (10 of 15). security.txt valid per the 30 September check, SOC 2 Type II, ISO 27001 and HIPAA claims on the pricing page, and a public SOC 3 report since 21 April 2026. No bug bounty found (17 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.

## 4. Agent ergonomics, 80 out of 100, up to 3.3 more on the total

Why it scored 80: The open-source server registers 61 tools by default (28 core and alerting, 15 vault, 19 browser interaction), all at once, and the docs list a smaller core of 10. There's a code path to mount interaction tools per session, but registration is always-on for clients that don't refetch. On the API side, format=markdown and the extraction API keep responses small (10 of 25). Output format, extraction templates, crawler limits and 100-URL batches (15 of 20). Error codes an agent can branch on, with retryable and billed flags (20). Tools carry readOnlyHint, destructiveHint, idempotentHint and openWorldHint, and mutating vault and alert tools need confirm=true after a preview (20). web_get_page works with a URL alone, and official SDKs exist for Python, TypeScript, Go and Rust (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.

## 5. Transparency & trust, 77 out of 100, up to 2 more on the total

Made of editorial 58, provenance 96.

Why it scored 77: Closed service with terms updated 2026-09-30 under US law. SDKs are BSD 2-Clause, and the MCP source is public with no LICENSE file and ISC in package.json (15 of 30). Privacy policy updated 14 May 2026 with a published DPA and HIPAA BAA, but retention is as long as necessary, and the retention of scraped content isn't stated (18 of 30). Release notes date deprecations and removals, for example Scrapium v153 dropping legacy network interception on 11 September. No standing deprecation policy found (15 of 20). Stripe, Google Analytics, Google Ads and Google Cloud are named, with SCC and Data Privacy Framework transfers, but there's no full subprocessor list (10 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):

- Domain age: scrapfly.io, registered 2018-09-12 (8 years) (11 of 15)

## 6. Maintenance & community, 84 out of 100, up to 1.4 more on the total

Why it scored 84: MCP v1.3.4 and both SDKs at 0.12.0 on 2026-09-08, and release notes dated 2026-09-25 (30). MCP tags v1.3.1, v1.3.2, v1.3.3 and v1.3.4 between 26 August and 8 September, and 25 commits to the MCP repo since 3 July (20). Dated release notes most weeks and a support address. We couldn't read GitHub issues, since the issue search is blocked to our reader (12 of 25). In the official MCP registry as io.scrapfly.mcp/mcp under the vendor's domain, and SDKs are current (15). MCP on Go 1.26 and go-sdk v1.7.0. Its CI runs only a reference check and the npm publish, while the TypeScript SDK runs tests in CI (7 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. Schema & documentation, 95 out of 100, up to 0.8 more on the total

Why it scored 95: A Swagger reference per the 30 September check, and the MCP server exposes it as a web_scraping_openapi_specification resource. Every MCP tool has JSON Schema inputs generated from typed Go structs (25). llms.txt and Markdown twins such as pricing.md per the 30 September check. Our reader was refused by robots.txt today (10). Descriptions say when to use a tool and when to switch, for example web_get_page falls back to web_scrape for tuning and to cloud_browser_open for interaction. A few are long and steer the model's next step (17 of 20). Typed inputs with enums, for example the vault service discriminator, and required fields (13 of 15). Error catalogue of 100+ ERR:: codes, each with HTTP status, retryable and billed flags and its own doc link (15). Dated release notes, newest 25 September 2026 (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.

## 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: llms.txt today. robots.txt refused our reader, so the llms.txt and Swagger facts are from the 30 September check
- The docs list docs_search and monitoring_get_metrics tools that aren't in the public MCP source, so the hosted server's tool count may differ from 61
- Whether Scrapfly keeps a status history anywhere. scrapfly.io/status shows none and status.scrapfly.io doesn't resolve
- unchecked: GitHub issue responsiveness on scrapfly-mcp

## Weaknesses

- The open-source MCP server registers 61 tools at once, a heavy tools/list
- The canonical API auth form and one MCP option put the key in the URL query string
- Status page shows an overall state only, with no components or incident history, and status.scrapfly.io doesn't resolve
- Residential requests cost 25 credits and per-target Unblocker surcharges aren't published
- No SLA, and the terms disclaim uptime

## 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 `X-Scrapfly-Api-Key` as a header rather than `key=` so the key stays out of logs
- Pass `unblocker=false` to the MCP web_scrape tool for plain pages, since it's on when omitted. `asp` is the deprecated alias
- Set `cost_budget` on each scrape so an Unblocker upgrade can't blow through credits
- Check `retryable` in the error body before retrying, and stop on `ERR::THROTTLE::MAX_API_CREDIT_BUDGET_EXCEEDED`
- Use web_get_page for a simple fetch and save cloud_browser_open for pages that need clicks

## What the review panel asked for

- publish Unblocker surcharges per target
- core tools by default

## 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.