linkedin-profile-optimizer

A Claude Skill / Claude Code plugin that audits a LinkedIn profile against a specific target job title, role, or designation and produces a decision-support Markdown document with multiple phrasing options per profile area — headline, About/summary, Skills, and each Experience entry — rather than a single prescriptive rewrite.

Useful for: LinkedIn optimization, resume-to-LinkedIn reconciliation, career pivots, promotion cases, personal branding, and recruiter-visibility improvements.

See SKILL.md for the full skill definition (trigger conditions, step-by-step process, output structure).

Inputs

• LinkedIn profile export (PDF) — required
• Target designation (e.g. "Staff Engineer", "VP Engineering") — required
• Resume (PDF) — optional; if provided, the skill reconciles it against LinkedIn and asks before trusting any metric or project that doesn't appear on LinkedIn itself

Output

A single .md file: outputs/[Name]_LinkedIn_Options_[Designation].md, containing a brief positioning read followed by 2–4 concrete options per profile area, each with a recommendation.

Install

This is an Agent Skill — a SKILL.md file with YAML frontmatter that Claude loads automatically based on its description. There's no separate "run" step to install; you place the folder where the surface you're using looks for skills, and Claude picks it up on its own once the request matches.

Claude Code — as a plugin (recommended)

This repo ships a .claude-plugin/plugin.json and .claude-plugin/marketplace.json, so it's installable as a plugin without manual file copying:

/plugin marketplace add nikhilvdev/linkedin-role-optimizer
/plugin install linkedin-profile-optimizer

This also means the plugin can be listed in any marketplace catalog (official or community) that indexes GitHub repos with a .claude-plugin/marketplace.json — see Optimizing for discovery below.

Claude Code — manual skill install (CLI / VS Code / JetBrains extensions)

Personal skills (available in every project):

mkdir -p ~/.claude/skills
git clone https://github.com/nikhilvdev/linkedin-role-optimizer.git ~/.claude/skills/linkedin-profile-optimizer

Project-only skill (checked into a specific repo, shared with teammates who pull it):

mkdir -p .claude/skills
git clone https://github.com/nikhilvdev/linkedin-role-optimizer.git .claude/skills/linkedin-profile-optimizer

Either way, the folder must contain SKILL.md at its root (it does). Restart Claude Code (or start a new session) so it re-scans the skills directories — no further setup needed.

claude.ai (web) and Claude Desktop

1. Go to Settings → Capabilities → Skills (web) or the equivalent Skills section in Desktop settings.
2. Upload this folder as a .zip, or point the "Add skill" flow at SKILL.md directly if a file picker is offered.
3. Enable the skill for the conversation/project where you want it available.

Claude API (Agent Skills via the Files/Skills API)

Upload SKILL.md (and this folder) through the Skills API and reference it in your agent_skills config when creating a run. See the Agent Skills docs for the exact request shape — the skill package itself (this repo) doesn't change between surfaces.

How to invoke it

This skill is not a slash command — you don't type /linkedin-profile-optimizer. Claude loads it automatically when your request matches the trigger described in SKILL.md's frontmatter. Just describe what you want in plain language and attach the PDF(s).

It will trigger on requests like:

• "Optimize my LinkedIn for a Staff Engineer role" (+ attach LinkedIn PDF export)
• "How does my profile look for a VP Engineering position?"
• "Help me reposition my LinkedIn for a Product Management pivot"
• "Review this LinkedIn export against the Senior Engineering Manager title"
• Attaching just a LinkedIn PDF with no resume and asking "what should I change to target [role]?"

What to provide:

1. Your LinkedIn profile export as a PDF (Me → Resources → Save to PDF on your own profile), attached to the message.
2. The exact target designation/title you're optimizing for — be specific ("Staff Engineer" not "a promotion").
3. (Optional but recommended) Your resume as a PDF, if it has more detail (metrics, projects) than what's currently on LinkedIn.

Example prompt:

Here's my LinkedIn export and resume. Optimize my profile for a "Head of Product" role — I'm currently a Senior PM and want to make the case for the jump.

If you only give a target designation with no PDF, Claude will ask for the LinkedIn export before proceeding — it won't work from a resume alone, since recruiters search against LinkedIn's actual live field values.

What you get back: a single Markdown file (outputs/[Name]_LinkedIn_Options_[Designation].md) delivered as an attachment/file, not pasted into chat — open it, pick the options that match your voice, and ask Claude to write the final copy for whichever ones you choose.

Optimizing for discovery

This repo is set up to surface in three separate places people "search for skills":

1. Claude's own auto-matching (inside Claude Code / claude.ai / Desktop) — driven entirely by SKILL.md's YAML description. It's written with the target job titles, task synonyms ("rewrite," "audit," "reposition," "career pivot," "promotion case"), and example phrasings spelled out, since Claude matches on that text, not the repo name or README.
2. Claude Code plugin/marketplace search — .claude-plugin/plugin.json and .claude-plugin/marketplace.json carry keywords, category, tags, description, and author fields, which is what /plugin marketplace add / marketplace browsing indexes on.
3. Plain GitHub/web search — driven by the repo's topics and README text, not the plugin manifests. Two things worth doing once, outside of Claude:
   • Set repo topics (Settings → topics, or gh repo edit --add-topic) to something like: claude, claude-code, claude-skill, claude-plugin, agent-skills, linkedin, resume, career, job-search.
   • If you want broader reach, submit the plugin to Anthropic's community marketplace (anthropics/claude-plugins-community) per the plugin docs — it goes through their validation/safety screening before listing.

License

See LICENSE.