Linkedin stats

Skill gokul-kulkarni/content-studio-plugin/skills/linkedin-stats

Turn LinkedIn analytics into the normalized record shape the analyze skill reads — post impressions, reach, reactions, comments, reposts, follower growth, and audience demographics — from the native analytics export you download. Use whenever someone asks how their LinkedIn posts are doing, wants impressions/reach numbers, asks which carousel performed best, or is gathering data for an outreach/quality analysis.From its SKILL.md

Install

npx -y skills add gokul-kulkarni/content-studio-plugin --skill linkedin-stats

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

One thing to look at

1 stars1 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.

SKILL.md

2.8 KB, as published. Nobody here has run it

LinkedIn Stats

Turn the numbers for $ARGUMENTS (a LinkedIn creator) into the normalized record shape analyze reads. LinkedIn gates analytics behind approved partner vendors and blocks unauthenticated access, so the native export is the source — you download it and drop it in; this skill parses it.

The export (XLSX)

Your profile → Analytics (creator view) → Export → downloads an XLSX. The aggregate export has five sheets, each worth reading:

DISCOVERY — total impressions + members reached for the window.
ENGAGEMENT — daily impressions + engagements (spot the launch inflection and best days).
TOP POSTS — two separate rankings side by side: top posts by engagements and top posts by impressions (URLs + dates). They rarely match — compute engagement rate (engagements ÷ impressions) to see which posts were genuinely strong vs merely widely shown.
FOLLOWERS — total followers + daily new-follower counts (map signups to publish days).
DEMOGRAPHICS — company / location / seniority / job title / industry of your audience. This is the richest ICP signal any platform gives you — pass it to analyze verbatim.

Parse the XLSX directly (it's a zip of sheet XML; read sharedStrings.xml + each sheetN.xml, or use a spreadsheet lib). For a single post, "View analytics" on the post shows impressions/reach/engagement — a paste works too.

⚠️ Reach ≠ right reach. LinkedIn's top-impressions post is often a personal/off-topic one that spread through your legacy network at a low engagement rate and the wrong audience. Rank by engagement rate, not raw impressions, so a viral-but-off-target post isn't mistaken for a content win. Flag such outliers for analyze rather than averaging them in.

What to produce

Emit the normalized records (schema in the analyze skill): one PostRecord per post with impressions, reactions, comments, shares (reposts) filled where available, format = carousel/post, plus an AccountSummary for follower count/growth. Note the source (export). The demographics sheet has no place in the per-post schema — carry it alongside as ICP context for analyze. Then hand off to analyze.

Privacy

Personal analytics — keep them local, never send to an external service.