# A founder’s 15-minute LinkedIn commenting routine with Gemini Source: https://finchreach.com/blog/gemini-linkedin-founder-comments Published 2026-10-28 by Finchreach team. Tags: gemini, gemini-cli, linkedin, mcp, founders, comments. Use Gemini to find feed posts worth commenting on and draft thoughtful comments you approve. Setup for the Gemini app and Gemini CLI. ## The short answer Founders get more from thoughtful comments on other people’s posts than from posting more often, but finding the right posts takes time. An assistant connected to your LinkedIn can read your feed, pick the posts where you have something real to add, and draft comments for you to approve. Finchreach provides that connection as a remote MCP server (MCP is a standard plug that lets AI apps use outside tools). Both the Gemini app at gemini.google.com and Gemini CLI can connect to Finchreach. The app uses a custom app with OAuth sign-in; Gemini CLI takes one command. The routine below uses Gemini CLI, and every comment waits in your [approval queue](https://finchreach.com/approvals) before it is posted. ## Can the Gemini app use my LinkedIn account? Yes, as of October 2026. The Gemini app lets you add custom apps under Settings → Connected Apps, using OAuth with dynamic client registration, which Finchreach supports. Add a custom app with the URL `https://mcp.finchreach.com/mcp`, then sign in to Finchreach, choose the LinkedIn account and scopes, and approve. Google limits the feature (18+, US, personal account, Keep Activity on, English). Source: [Gemini Help](https://support.google.com/gemini/answer/17209137). Your options: - The Gemini app, with a custom app and OAuth sign-in, if you meet Google’s requirements. - Gemini CLI, Google’s open-source terminal agent, which supports remote MCP servers with OAuth or custom headers. Covered below and in the [Gemini CLI guide](https://finchreach.com/guides/gemini-cli). - Any other MCP client, such as [Claude Code](https://finchreach.com/guides/claude-code) or [Cursor](https://finchreach.com/guides/cursor). One more caveat: a third-party report, which we have not been able to verify, says Gemini CLI stopped serving consumer AI Pro and Ultra accounts in June 2026. If your sign-in no longer works, check Google’s current documentation for the recommended CLI. ## Why should founders comment on LinkedIn posts? A good comment on a post by a customer, an investor or a respected operator puts your name in front of their audience, in a context where you are adding something. It also builds relationships with the people you comment on, which a post of your own rarely does. The hard part is not writing; it is triage. Your feed is long, and most posts are not worth a comment from you. An assistant can do that filtering: - Read your feed with `linkedin_get_feed`, sorted by recent or by relevance. - Find posts on topics you know deeply, from people you care about. - Skip posts where you have nothing specific to add. - Draft comments that add a data point, a story or a respectful disagreement, not “Great post!”. ## What does the 15-minute routine look like? 1. Open Gemini CLI and ask it to read your recent feed and list candidate posts with a reason for each. 2. Pick three to five. Tell it what you actually think about each one, in a sentence or two. 3. It turns your notes into comments, keeping your point and your voice, well under the 1,250-character limit. 4. It queues each comment with `linkedin_comment_on_post`, and optionally a reaction with `linkedin_react_to_post`. Each returns `pending_approval`. 5. You approve them in Finchreach, ideally spread through the morning rather than all at once. Step 2 matters most. Comments that come from your real opinion read like you. Comments generated from nothing read like everyone else’s. ## Which prompts work? > Ask your assistant: “Read my LinkedIn feed, sorted by recent, two pages. I run a B2B payments startup. List up to 8 posts where a founder in my space could add something specific: author, their headline, the post’s point in one line, and why I might comment. Skip announcements and engagement bait.” > Ask your assistant: “For post 2, my take is: we tried usage-based pricing for a year and churn went up with small customers. For post 5: I disagree, onboarding calls still matter at our price point. Turn these into comments under 70 words each, first person, no hashtags, no flattery opener.” > Ask your assistant: “Queue those two comments with linkedin_comment_on_post and like post 2. Give me the action ids.” Save your positioning, voice and topics in a `GEMINI.md` file in the folder where you run the CLI, so every session starts with the same context. ## How do I connect Gemini CLI to LinkedIn? 1. [Create a Finchreach account](https://finchreach.com/sign-up) and connect your LinkedIn account on the Accounts page. 2. Create an API key on the API keys page and export it as `FINCH_KEY`. 3. Run the command below. Add `-s user` to store it in `~/.gemini/settings.json` for every folder instead of the current project. 4. Run `/mcp` in Gemini CLI to see the Finchreach tools. 5. Details: the [Gemini guide](https://finchreach.com/guides/gemini), the [Gemini CLI guide](https://finchreach.com/guides/gemini-cli) and the [docs](https://finchreach.com/docs). Add Finchreach to Gemini CLI: ```bash gemini mcp add --transport http --header "Authorization: Bearer $FINCH_KEY" \ finchreach https://mcp.finchreach.com/mcp ``` Source: [Gemini CLI MCP docs](https://geminicli.com/docs/tools/mcp-server/), as of October 2026. ## Is it risky to comment on LinkedIn with an AI assistant? > Automating a personal LinkedIn account carries risk under LinkedIn’s User Agreement. Finchreach’s approval queue and a human pace lower that risk; they do not remove it. The reputational risk is just as real. People notice generic AI comments, and a founder’s name is the company’s name. Keep the volume low, keep the opinions yours, and reject any draft you would not have written. ## Frequently asked questions Can Gemini delete a comment if I change my mind? Yes, `linkedin_delete_comment` removes it, after approval. Can it reply to comments on my own posts? Not as threaded replies. It can check your notifications for new comments and add a top-level comment on your post that responds, after approval. Does the Gemini app need an API key? No. It connects with OAuth: you sign in to Finchreach and approve. The [Gemini guide](https://finchreach.com/guides/gemini) has the details. Related reading: [event follow-ups with Gemini CLI](https://finchreach.com/blog/gemini-cli-linkedin-event-follow-up) and [what a LinkedIn MCP server is](https://finchreach.com/blog/what-is-a-linkedin-mcp-server).