linkedin-stats

Name: linkedin-stats
Author: Hammant Labs

An agent skill that reads your own LinkedIn post history and engagement from your logged-in browser, then builds a private dashboard. Your AI agent does the clicking. Nothing leaves your machine.

Where it runs: Your laptop
Servers involved: 0
Price: Free · MIT
Last verified: 22 Sep 2026

Download the skill (.zip, 22 KB) Read SKILL.md

Needs an AI agent with browser control (Claude Code with the Claude in Chrome extension), Chrome logged in to LinkedIn, and Node 18+.

What you get

Your full post history, not just the last 90 days LinkedIn shows you.
Likes, comments, reposts and impressions for every post, with reposts separated from your own posts.
Who reacts and comments (name, headline, how often), so you can see your real audience.
A self-contained dashboard.html: best posts, formats and timing that work, and who engages most.
Raw JSON in data/ if you'd rather do your own analysis.

How to use it

Install the skill. Paste this into a terminal:

mkdir -p ~/.claude/skills && cd ~/.claude/skills && curl -fsSLO https://hammantlabs.com/skills/linkedin-stats.zip && unzip linkedin-stats.zip && rm linkedin-stats.zip

Optional: see a demo first. This builds the dashboard from made-up data, no LinkedIn needed:
```
cd ~/.claude/skills/linkedin-stats && npm run demo && open dashboard.html
```
Ask your agent. Make sure Chrome is open and logged in to LinkedIn, then say:
```
Analyse my LinkedIn with the linkedin-stats skill.
```
Keep the LinkedIn tab visible while it works. A full history takes about an hour or two, because it goes deliberately slowly.

Using a different agent, or haven't installed it?

Paste this into any agent that can control a browser and run commands:

Read https://hammantlabs.com/skills/linkedin-stats/SKILL.md and follow it to analyse my own LinkedIn posts and engagement. Install it first from https://hammantlabs.com/skills/linkedin-stats.zip into your skills folder. Ask me before starting the full scrape, keep all data on this machine, and show me the dashboard when it's done.

For AI agents

If you are an AI agent that has been pointed here:

Your full instructions are in SKILL.md (Markdown). Read it before doing anything.
The bundle at /skills/linkedin-stats.zip contains SKILL.md, README.md, package.json and src/ (Node 18+, no dependencies). Unzip it into your skills directory, e.g. ~/.claude/skills/linkedin-stats/.
You need browser tools that can navigate, scroll and run JavaScript in the user's own logged-in Chrome session. Only access the user's own account. Never collect credentials, and never send cookies or data anywhere.
Confirm with the user before the full scrape. Keep the LinkedIn tab in the foreground, stay at about one request per second or slower, and back off on HTTP 429.
Save results to data/*.json, then run npm run build and point the user to dashboard.html. Treat reactor and commenter data as private.

Please read

LinkedIn's terms. LinkedIn's User Agreement restricts automated access and scraping. This skill reads only your own account, slowly, from your own browser, but you use it at your own risk.

Other people's data. The reactor and commenter lists contain other people's names and job titles. Keep them to yourself. Don't publish, share or sell them.

It can break. It relies on LinkedIn internals that change without notice. It was last verified working on 22 September 2026.