linkedin-mcp
Server Configuration
Describes the environment variables required to run the server.
| Name | Required | Description | Default |
|---|---|---|---|
| LINKEDIN_DB_PATH | No | Scheduler database. | $LINKEDIN_MCP_HOME/schedule.sqlite |
| LINKEDIN_POLL_MS | No | Scheduler poll interval. | 30000 |
| LINKEDIN_HEADLESS | No | Run non-login actions headless. Login is always headed. | false |
| LINKEDIN_MCP_HOME | No | Root for session + scheduler state. | ~/.linkedin-mcp |
| LINKEDIN_PACING_MS | No | Randomized delay between UI steps (`min,max`). | 400,1200 |
| LINKEDIN_SESSION_DIR | No | Persistent Chrome profile (your logged-in session). | $LINKEDIN_MCP_HOME/session |
| LINKEDIN_LOGIN_TIMEOUT_SEC | No | How long login waits for you. | 300 |
Instructions
Guidance the server publishes about itself, which clients place ahead of the tool catalog so the model reads it before choosing anything.
This server publishes no instructions, or was last inspected before Glama recorded them.
Capabilities
Features and capabilities supported by this server
Protocol revision2025-11-25
| Capability | Details |
|---|---|
| tools | {
"listChanged": true
} |
Tools
Functions exposed to the LLM to take actions
| Name | Description |
|---|---|
| linkedin_loginA | Open a real Chromium window and wait for you to log in manually (including 2FA / security checkpoints). The session is saved locally and reused; your password is never read or stored. Run once, or again if the session expires. |
| linkedin_session_statusA | Report whether the saved LinkedIn session is still logged in, and warn if core page selectors have drifted (a maintenance signal). |
| create_postA | Publish a text post to your personal LinkedIn profile, now. For a future time use schedule_post (which first asks the user native vs local scheduler). |
| schedule_postA | Publish a post at a future time, through ONE of two schedulers. There is no default — ask the user which one they want BEFORE calling this, and state the trade-off:
|
| list_scheduled_postsA | List queued/scheduled posts and their status. LOCAL queue only — posts handed to LinkedIn's native scheduler are held by LinkedIn and only visible on linkedin.com. |
| cancel_scheduled_postA | Cancel a still-pending post in the LOCAL queue by id. Natively scheduled posts must be canceled in LinkedIn's own scheduled-posts view. |
| list_my_postsA | List your recent posts (their refs can be passed to read_post_comments). |
| read_post_commentsA | Read comments on a post. |
| reply_to_commentA | Reply to a comment (use commentRef from read_post_comments). |
| list_pagesA | List Company/showcase Pages surfaced to your account. Best-effort — may include followed (non-admin) companies; verify admin rights before posting as a Page. |
| create_page_postA | Publish a text post AS a Company/showcase Page you admin. |
| read_feedA | Read a light snapshot of your home feed (best-effort; feed layout changes often). |
Prompts
Interactive templates invoked by user choice
| Name | Description |
|---|---|
No prompts | |
Resources
Contextual data attached and managed by the client
| Name | Description |
|---|---|
No resources | |
TDQS
Scored across 12 tools
Each tool targets a distinct action/resource: authentication, session status, immediate post, scheduled post, queue management, post listing, comment reading/reply, page listing/page post, and feed reading. No two tools overlap in purpose.
Most tools follow a clear verb_noun pattern (create_post, schedule_post, list_scheduled_posts, cancel_scheduled_post, list_my_posts, read_post_comments, reply_to_comment, list_pages, create_page_post, read_feed). The exceptions are linkedin_login and linkedin_session_status, which use a prefix style that deviates from the rest.
12 tools cover the core LinkedIn automation domain—auth, posting (immediate and scheduled), scheduling queue management, comments, pages, and feed—without excess. This is a well-scoped count for the server's purpose.
The posting lifecycle is well covered (create, schedule, list, cancel), along with comments and page posts. Missing operations like editing or deleting posts are minor gaps; agents can work around them, but they are noteworthy for a complete LinkedIn content management surface.