Automate LinkedIn With AI Agent Skills

Gives your AI agent the ability to run LinkedIn operations end-to-end. One skill handles general automation (searching, messaging, managing connections, posting), while the other runs a managed lead pipeline with qualification and scheduled invites. Together they let agents execute LinkedIn workflows without manual tool-switching or state management.

Once set up with Node 20+, Linked API tokens, and linkedin-cli, the skills are available to your agent. For general work, pass a command (fetch profiles, send messages, search people/companies, manage connections, create posts, run custom workflows). For lead pipeline work, import leads from a search, your agent qualifies them against your ICP, the skill stores them in a local database, then a background scheduler sends invites on a per-account schedule and cleans up stale pending requests. Output: LinkedIn operations executed via real browser, results in JSON.

LinkedIn operations run through your agent without leaving your workflow, and a lead pipeline scales invitation pace across multiple accounts on an automatic schedule.

Requires Node.js ≥ 20 because the skills use modern async patterns and the better-sqlite3 module. Requires @linkedapi/linkedin-cli globally installed; it handles LinkedIn interaction via a real cloud browser. Linked API tokens (available at app.linkedapi.io) authenticate both the CLI and the skill. For linkedin-growth, npm install --omit=dev pulls in production dependencies (better-sqlite3, native module build required).

linkedin-cli requires Node.js ≥ 20 and Linked API tokens (requires account at app.linkedapi.io). linkedin-growth requires SQLite and local database setup; all leads and state are local only. Real browser operations (via linkedin-cli) take 30 seconds to several minutes per action. LinkedIn's own rate limits and account restrictions apply. Sales Navigator search support requires an active Sales Navigator subscription.