jevcrawl

# Fix list: jevcrawl

From Anchor Terminal's listing at https://www.anchorterminal.com/tools/jevcrawl, the October 2026 research run, assessed 15 October 2026. Grade B, 63.7 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 jevcrawl: 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, 30 out of 100, up to 8.8 more on the total

Why it scored 30: No payment protocol (0). The package is free, so read with the rule for the paid option it needs, TypeSafe's Jev. Per-token price published at $0.042 per million input tokens (20). No free Jev tier found. `scrape` without a goal works with no key, which is a small free path (5 of 20). A person has to sign in at console.typesafe.ai, and TypeSafe describes Jev as early access. The console answered 403 to our reader. Keyless goal-less scraping earns a little (5 of 20).

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, 59 out of 100, up to 7.2 more on the total

Why it scored 59: Read with the API and MCP lines, since the server runs locally but sends data to a hosted API. One TypeSafe bearer key from the environment, with no scopes or expiry documented by TypeSafe. The README says the key is never written to disk or logs (20). Every tool only reads, a crawl stays on the start host, and robots.txt is followed by default. There is no check against private, loopback or link-local addresses, so an agent can be steered to internal hosts (12). Pages are untrusted content. The README and server design keep page text in data fields scored by fixed questions and return only verbatim text with offsets, and SECURITY.md names a change in fetch behaviour through page content as in scope. These are the author's claims, untested by us (12). Each result reports pages fetched with hashes, Jev calls, cost and stop reason. No persistent call log (7). SECURITY.md with GitHub private reporting and a one-week answer. No bounty, no advisories yet, and TypeSafe publishes no security.txt (8).

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. Reliability, 68 out of 100, up to 6.4 more on the total

Why it scored 68: Read as local software, though `crawl` and `map` also depend on TypeSafe's hosted Jev API. Installs from PyPI with Python 3.11 or later stated (20). CI runs ruff, mypy strict and 174 offline tests on Python 3.11 and 3.13, and every CI run on main since 11 October 2026 passed (25). Eight issues, all opened by the maintainer as a roadmap, two closed. No outside bug reports, but the project is three days old, so there is little record to read (15). Semver tags with GitHub release notes, no CHANGELOG file, v0.1.0's notes only a commit link, and breaking changes not marked (8). Version 0.2.0, classified Development Status 3 Alpha (0).

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.

## 4. Transparency & trust, 45 out of 100, up to 4.8 more on the total

Made of editorial 63, provenance 27.

Why it scored 45: Apache-2.0 in LICENSE, pyproject.toml and as detected by GitHub (30). The README says plainly that page text, link labels and URLs go to TypeSafe and that answers are cached locally. TypeSafe's policies agree on no training but give no retention period, and jevcrawl has no privacy statement of its own (15). No deprecation policy. The model version is pinned and changes are in release notes (3). No telemetry or analytics found in the source. Outbound traffic is the target site and api.typesafe.ai, both stated in the README, so there is nothing hidden to opt out of, though no explicit telemetry statement exists (15).

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):

- Legal entity named: not found (0 of 20)
- Domain age: github.com/RayenMlayeh, no registry record we could read (0 of 15)
- Status page: not found (0 of 10)
- security.txt: not found (0 of 10)

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

Why it scored 80: Three compact tools (25). `max_pages`, `max_seconds`, `return_tokens` and `limit` bound every call, and `scrape` can return a whole page or only the passages for a goal. No pagination is needed (18). Expected failures come back as `{error, message}` with a message that says what to do, and `stopped_because`, `degraded` and `needs_js_urls` tell the agent why a result is thin (19). All tools read only and repeats are answered from a local cache at no cost, but there are no readOnlyHint or idempotentHint annotations (10). Only `goal` and `start_url` are required, with measured defaults. The Python library is usable but undocumented (an open issue), and there is one language (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.

## 6. Schema & documentation, 81 out of 100, up to 3.1 more on the total

Why it scored 81: Typed JSON Schema inputs on all three tools, generated from the Python signatures, with required parameters marked (25). No llms.txt. The repository ships an agent skill, `skills/jevcrawl/SKILL.md`, as Markdown for agents (5). Tool descriptions and server instructions say when to use `crawl`, `scrape` and `map`, how to phrase a goal and how to read each stop reason (18). Integer and number types with defaults, but the minimums the code enforces are not in the schema and `stopped_because` is documented as prose, not an enum (10). A real result in the README, and error codes listed in the skill with what each means (13). Semver releases with notes, and the Jev model version pinned per release (10).

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, 68 out of 100, up to 2.8 more on the total

Why it scored 68: v0.2.0 on 15 October 2026 (30). Three releases between 13 and 15 October 2026 (20). No outside issues or pull requests yet. The maintainer's own roadmap issues were closed by releases within two days (8 of 25). Not in the official MCP registry, which returns nothing for jevcrawl (0). Current dependencies, a uv.lock, passing CI and PyPI trusted publishing (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.

## 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.

- Judgement call. Listed in web-scrapers, as the lead suggested, because its job is crawling and extracting passages from websites. It has none of the proxy or anti-bot capabilities of the category's hosted APIs, so `web.crawl` comes first.
- The lead was right. The repository's first commit in the current history is dated 13 October 2026 with v0.1.0 launch preparation the same day. GitHub records the repository as created on 11 October 2026 and CI runs from 11 and 12 October, so earlier commits were rewritten. `firstReleased` is v0.1.0 on PyPI and GitHub, 13 October 2026. The project is three days old.
- We did not install or run the package or call Jev. Every claim about behaviour comes from the source, the tests and the README. The benchmark figures are the author's.
- unchecked: whether TypeSafe's console accepts new sign-ups or keeps a waitlist. console.typesafe.ai answered 403 to our reader, and typesafe.ai says Jev is in early access.
- jevcrawl's README calls it safe against prompt injection by design. The design limits page text to data fields, but Jev's own limitations page says injected text can move a score, which the README also says.
- No affiliation between Rayen Mlayeh and TypeSafe AI was found. jevcrawl uses the TypeSafe name and console link in its docs.

## Weaknesses

- First released on 13 October 2026, version 0.2.0 and marked alpha, with one maintainer
- `crawl` and `map` need a paid TypeSafe key, and TypeSafe describes Jev as early access with no free tier found
- Page text from every crawl goes to TypeSafe, whose policies give no retention period
- No check against private, loopback or link-local addresses, so an agent can point it at internal hosts
- No MCP tool annotations, no llms.txt and no entry 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.

- Phrase `goal` as one specific question and start with `max_pages` between 10 and 30
- Check `stats.stopped_because` and `stats.degraded`, and read `uncertain` before reporting that something is missing
- Use `scrape` without a goal when you need a whole page, since it makes no paid Jev call and needs no key
- Handle `{"error": code, "message": ...}` results, for example `auth_error` means `TYPESAFE_API_KEY` is missing or rejected
- Pages in `stats.needs_js_urls` were not read. Tell the user rather than guessing their content

## 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.