Hermoso is a marketing MCP server. Connect LinkedIn once and it becomes callable from Claude, ChatGPT, Claude Code, Cursor, Codex, Cline or the CLI — the same connection the web app uses, so nothing is duplicated and nothing is re-authorised per client.
What Hermoso does with LinkedIn
Post to a company Page or your profile, run ad campaigns, and answer comments on your Page’s posts. Your profile is off unless you turn it on.
Publishing: Publishing · Company Pages · Ad campaigns · Comments
Driving it from an agent
The LinkedIn tools an agent can call, by name (28 today):
associate_linkedin_conversion_campaigns create_linkedin_ads_campaign create_linkedin_ads_campaign_group create_linkedin_ads_creative create_linkedin_conversion_rule create_linkedin_event delete_linkedin_ads_object delete_linkedin_comment linkedin_ads_report linkedin_audience_count linkedin_bid_pricing linkedin_page_analytics list_linkedin_ads_campaigns list_linkedin_comments list_linkedin_conversion_rules list_linkedin_events list_linkedin_pages manage_linkedin_post post_to_linkedin post_to_linkedin_page reply_to_linkedin_comment search_linkedin_ad_library search_linkedin_ads_targeting search_linkedin_brand_partnerships send_linkedin_conversions set_linkedin_ads_budget set_linkedin_ads_status update_linkedin_conversion_rule
Connecting
Open the app, go to Workspace ▸ Connectors and connect LinkedIn. One authorisation covers the web app, the MCP server and the CLI. See MCP & CLI for the endpoint and the client setup, or every channel we connect to.
Frequently asked questions
Can an AI agent post to LinkedIn through Hermoso?
Yes. The LinkedIn surface is exposed over one OAuth'd MCP endpoint, so Claude, ChatGPT, Cursor, Codex, Cline and the CLI can all drive it once the account is connected.
Do I need a separate LinkedIn connection for the app and for MCP?
No. One authorisation covers the web app, the MCP server and the CLI — it is the same stored connection.