test(crawl): add a first-install acceptance test for the crawl command

[StefanoGuerrini]

Stacked on #31 (feat/crawl-command-v2).

Summary

This PR adds an acceptance test for the zenrows crawl command of @zenrows/cli. It installs the release artifact the way a new user does and runs every Crawl feature against the live API. It stays in its own PR, so the feature PR stays focused.

What it does

acceptance/crawl/run.sh builds the release artifact from this checkout: npm install -g of the npm pack tarball. It installs the artifact in a clean container with digest-pinned images. The container gets no source, no build caches and no tokens, and it gets the key by name only. The run follows the read-me (README) Crawl quickstart, then runs checks 1 to 21 and the CLI extras. Each check writes one JSON result with status PASS, FAIL or SKIP. A markdown report lands in acceptance/crawl/out/, which git ignores. The files list in package.json keeps the harness out of the published package.

#	Feature	#	Feature
1	Create	12	Not found
2	Create with html	13	Invalid input
3	Idempotency	14	Gated option
4	Wait	15	Bad key
5	Wait runs out	16	Errors carry the server code
6	Get	17	Naming
7	List	18	Copy
8	Results	19	Failed crawl
9	Content	20	Results after stop
10	Download	21	Documented defaults
11	Stop

The CLI extras check that every subcommand prints JSON with --json. They check that Ctrl-C during crawl wait prints the resume hint and exits 130. They also check that the CLI rejects --wait and --html as unknown flags. acceptance/crawl/README.md gives the expected result of each check, the pinned versions, and how to read the report.

Three checks changed after the first commit:

• Check 14 sends --output-format json. A refusal before the API (INVALID_USAGE) now counts as a pass, as does a 422 (CRAWL_INVALID_REQUEST).
• Check 9 needs a fetched page. The stop in check 20 can land before the crawl fetches a page. Then check 9 starts one more small html crawl and waits for its end.
• The naming check (17) now also covers content_id.

Cost and safety

Every crawl uses depth 1 and max_pages 10 or less. A run starts five crawls, and check 9 can start one more. A run has at most 2 crawls active at a time, because an account has 3 active jobs, shared with Batch. If a create gets a 429, the run waits Retry-After and tries again. A clean-up step stops every crawl the run started, also when a check fails or on Ctrl-C. The script never prints the key. Before it ends, it scans every file under out/ for the key.

How to run it

The host needs bash, git and Docker, and a key with Crawl access.

export ZENROWS_API_KEY="$(cat path/to/key-file)"
bash acceptance/crawl/run.sh              # every check
bash acceptance/crawl/run.sh --no-key     # offline checks only; live checks SKIP
bash acceptance/crawl/run.sh --report     # rebuild the report from out/
REF=v1.2.3 bash acceptance/crawl/run.sh   # test a tag, branch or SHA from GitHub

The report is acceptance/crawl/out/crawl-first-install-report.md. Its verdict is "ready" when it has no blocker or major finding.

Squash-merge with the title as the message: the branch commits carry trailer lines that should not land.

[StefanoGuerrini]

Merged into the feature branch with a merge commit. It ships to main with #31.