feat(linkedin): summarize post counters

[gauravsaxena1997]

What and why

Once LinkedIn posts can be read from the browser session, agents often need a compact summary of the visible counters across recent posts. That summary should stay raw and auditable instead of adding custom ranking, scoring, or classification that may not match user intent.

This PR adds a read-only raw post analytics command built on the same visible post extraction path.

What this PR does

Adds a Strategy.COOKIE LinkedIn browser adapter:

opencli linkedin post-analytics --limit 5 -f json

Implementation details:

• Adds shared LinkedIn helpers for auth-wall checks, safe URL validation, limit parsing, text normalization, and evaluate-result unwrapping
• Adds posts-core.js for shared post extraction without registering an extra command
• post-analytics returns raw aggregate counters only: posts analyzed, total reactions, total comments, total reposts, total impressions, posts with media, posts with URLs, and latest-post counters
• Deliberately avoids custom engagement scores, topic labels, format labels, or positioning-gap recommendations
• Supports --limit from 1 to 100

Live verification

Verified against a logged-in LinkedIn profile session:

opencli linkedin post-analytics --limit 5 -f json

Returned raw aggregate counters for 5 visible posts, including total reactions, comments, reposts, impressions, and latest-post counters. No custom score or classification was included.

Checklist

• npm run build
• npx vitest run --project adapter clis/linkedin/post-analytics.test.js
• npm run typecheck
• node scripts/check-typed-error-lint.mjs
• node scripts/check-silent-column-drop.mjs

[jackwener]

Closed as superseded by consolidated PR #1731, merged as b6965a5.

[jackwener]

Superseded by consolidated PR #1731 (merged as b6965a5).