← All Skills
💼
LinkedIn Content Agent Skills
A pack of eleven Claude Skills for LinkedIn posts, comments, replies, DMs, profile audits, weekly planning, carousels, and repurposing, plus local Python rules for text cleanup and heuristic checks.
Key Features
- Post, comment, reply, and DM workflows
- 100-point profile audit and weekly plan
- Twenty-one hook formulas and carousel copy
- Local cleanup and five heuristic scores
Installation
In Claude Code, clone the repository and copy `skills/li-*` into `~/.claude/skills/`, or run `/plugin marketplace add Jakeschincariol/linkedin-agent-skill` followed by `/plugin install linkedin-agent`. Build `~/.claude/linkedin/voice.md` from templates/voice.md for better voice matching. The two li-human scripts require Python 3 and have no third-party dependencies.
How to Use
Call li-post, li-comment, li-reply, li-dm, li-profile, li-plan, li-carousel, li-repurpose, li-inbox, or li-audit for the matching task. Run li-human before showing copy; its scripts automatically handle characters, punctuation, and a word list, while structural issues need a human rewrite. Review every result, then paste or publish it manually.
Requirements
MIT and free, with no LinkedIn API, signup, or extra API key. It does not log in, scrape, or publish automatically. Posts, messages, profile details, and analytics pasted into Claude follow the chosen model provider's data boundary. voice.md, plan.md, and an optional log.md live under `~/.claude/linkedin/`. The li-human scripts only read local text and their lexicon and make no network calls. The human score is a local heuristic, not a result from any third-party detector and not a guarantee that content will evade AI detection. Verify each DM and comment to avoid impersonation or bulk spam.
Related Skills