chore: remove pagerangers-seo skill

[tripleawwy]

Why

netresearch/pagerangers-skill is archived and no longer maintained, so the catalog entry pointed at a dead skill. This removes it, mirroring feat: add pagerangers-seo skill which introduced it.

What

pagerangers-seo was referenced in five places, all removed:

File	Change
.claude-plugin/marketplace.json	plugin entry
README.md	catalog row
site/src/_data/descriptions_de.json	German description
site/src/_data/displayNames.json	display name
site/src/_data/groups.js	workflow group membership

Skill count in the README badge and intro line goes 40 → 39.

Also included: refreshed Playwright visual baselines. Removing a card makes both landing pages exactly 365px shorter, which failed visual-regression. The PNGs were regenerated on a CI-matching runner via refresh-visual-snapshots.yml (run 31379878153) rather than locally, since the baselines are environment-specific to chromium-desktop-linux.

Verification

scripts/validate.sh:

✓ Every plugin has a README catalog row

Marketplace valid: 39 plugins

site/scripts/check-categories.js:

Category check: 39/39 skills use canonical categories

No references remain:

grep -ril "pagerangers" . --exclude-dir=.git
(none)

Full Pages run green on the refreshed baselines — build, visual-regression and lighthouse all pass.

Notes

• check-orphans.js reports typo3-site-conformance and jujutsu-workflow missing relatedSkills. Verified pre-existing by running the same check on main — byte-identical output. Left untouched as out of scope.
• relatedSkills is sourced at build time via fetch:readmes, so a stale pagerangers-seo reference in another skill's README would surface in the Pages build rather than in this diff.

[github-actions]

Dependency Review

✅ No vulnerabilities or license issues or OpenSSF Scorecard issues found.

Scanned Files

None

[tripleawwy]

Copilot review not on this head

The copilot_code_review ruleset is unsatisfied and Copilot has posted nothing here: both reviews[] and requested_reviewers[] are empty.

The reviewer request was issued twice against the documented endpoint. Both calls returned HTTP 200 with the full PR object and requested_reviewers stayed empty — a silent no-op, most likely because the requesting token has WRITE rather than admin:

gh api repos/netresearch/claude-code-marketplace/pulls/93/requested_reviewers \
  -X POST -f "reviewers[]=copilot-pull-request-reviewer[bot]"
# HTTP 200, "requested_reviewers":[]

Not re-requesting further, per git-workflow's merge-gate guidance. Requesting the review from the web UI, or from an account with admin, is the remaining path.

Self-review

This is the author reviewing their own diff and is not a substitute for an independent review. Recording what was actually checked:

• All five pagerangers-seo reference sites removed; grep -ril pagerangers . --exclude-dir=.git returns nothing.
• scripts/validate.sh: 39 plugins, every plugin has a README catalog row.
• site/scripts/check-categories.js: 39/39 skills use canonical categories.
• site/scripts/check-orphans.js: the two reported orphans (typo3-site-conformance, jujutsu-workflow) verified pre-existing by running the same check on main — byte-identical output. Not touched.
• Both hardcoded skill counts updated 40 → 39: the README badge and the intro line.

visual-regression baselines

Removing a card makes both landing pages exactly 365px shorter, which invalidated the Playwright baselines:

Expected an image 1280px by 11869px, received 1280px by 11504px.  Snapshot: landing-de.png
Expected an image 1280px by 11774px, received 1280px by 11409px.  Snapshot: landing-en.png

They were regenerated on a CI-matching runner via refresh-visual-snapshots.yml (run 31379878153) rather than locally, since the PNGs are environment-specific to chromium-desktop-linux. Included in this PR; the full Pages run is green.

This does not clear the gate

Ruleset 15714055 has bypass_actors: null, so no note or self-review satisfies copilot_code_review. Merging still needs a Copilot review on b687e038 or a deliberate org-owner decision.

[sonarqubecloud]

Quality Gate passed

Issues
0 New issues
0 Accepted issues

Measures
0 Security Hotspots
0.0% Coverage on New Code
0.0% Duplication on New Code

See analysis details on SonarQube Cloud