Skip to main content
Glama
hermoso-ai

Hermoso

Official

Publish to a LinkedIn company Page

post_to_linkedin_page

Publish posts to a LinkedIn company page with text, image, video, carousel, or link previews. Supports alt text, captions, and thumbnails. Requires explicit user approval before immediate public publishing.

Instructions

Publish a post to one of the user’s LinkedIn COMPANY PAGES — text, plus optionally an image, a video, a 2–20 image CAROUSEL (LinkedIn calls it a MultiImage post; pass the slides in order as imageUrls[]), or a LINK POST with a real preview card (linkUrl). USE linkUrl WHENEVER THE POINT OF THE POST IS A LINK: LinkedIn disables URL scraping for API partners, so a url sitting in the text renders as plain text with no card, and the card’s title, description and image only exist if you pass linkTitle / linkDescription / linkThumbnailUrl — read them off the page and supply them. The media need not be a Hermoso render — it must be Hermoso-HOSTED because we upload the bytes to LinkedIn ourselves, and upload_file turns ANY file the user already has into such a URL. ORGANIC CAROUSELS ARE COMPANY-PAGE ONLY — a personal profile cannot publish one and is refused by name, so send a deck here rather than to post_to_linkedin. This is a DIFFERENT thing from post_to_linkedin, which publishes to the person’s own profile: pick the one the user actually asked for and never substitute. organizationId comes from list_linkedin_pages; omit it only when the account administers exactly one Page. This PUBLISHES immediately and PUBLICLY — ALWAYS show the user the exact text and get an explicit yes BEFORE calling. A VIDEO POST CAN CARRY CAPTIONS AND ITS OWN COVER, and both are attached only during the upload: pass captionsSrt (SubRip content — LinkedIn is watched with the sound off) and videoThumbnailUrl (otherwise LinkedIn picks a frame for you). LinkedIn does NOT allow the image, video, captions or thumbnail of a published post to be swapped afterwards, so get all of that right first (the copy can still be edited with manage_linkedin_post).

Input Schema

TableJSON Schema
NameRequiredDescriptionDefault
hookNoWHAT ANGLE THIS POST IS BUILT ON — the single most valuable field here, and the only moment it can ever be recorded. post_performance groups on it to answer "which hooks work", and it needs 5 posts sharing ONE hook before it will call anything a winner, so REUSE THE SAME WORDING across a campaign instead of rephrasing it every time. Best of all, pass a hook id from list_hooks (e.g. "direct_callout", "mid_problem", "before_after") — those fold onto a stable key however they are spelled, so a whole brand accumulates evidence on one row. Your own wording is fine too; it just only groups when you repeat it exactly. Omitting it means this post can never vote on which hook works.
textYesthe post text
titleNovideo title
altTextNoaccessibility alt text (max 4086 characters, ~120 recommended). A STRING describes every image; an ARRAY describes each slide of a multi-image post separately, in slide order — LinkedIn stores altText per image, and their own sample request carries a different one on each. Not available on a PERSONAL-profile post: LinkedIn’s member posting API has no alt-text field at all.
linkUrlNopublish a LINK POST — LinkedIn renders a real preview card for this URL instead of leaving a bare link in the text. Mutually exclusive with imageUrl / videoUrl / imageUrls: LinkedIn’s content field is a union, so combining them is refused by name rather than one being dropped.
subjectNoWHAT THIS POST IS ABOUT — the product, feature, offer or theme (e.g. "winter coat", "free trial", "founder story"). The second grouping axis in post_performance. Same rule as hook: reuse the exact wording so posts about one subject land in one group.
imageUrlNoa Hermoso-hosted image URL — a render (list_library), or ANY image of the user’s own passed through upload_file first. An arbitrary external host is refused.
videoUrlNoa Hermoso-hosted video URL — a render, or the user’s own footage via upload_file. LinkedIn processes it before publishing, which takes a minute.
imageUrlsNoCAROUSEL — an ORDERED list of image (and, where the channel allows, video) URLs published as ONE post the viewer swipes through. THIS IS NOT “post several” — it is a single post with several slides, which is what a multi-slide creative (a listicle, a “1/6 · SWIPE” deck) actually needs; publishing only its first slide tells the viewer to swipe at something that cannot. The ORDER is the product. Limits per channel: Instagram 2–10 (images, videos or a mix), Threads 2–20 (mix allowed), Facebook 2+ (Meta publishes no documented maximum; Hermoso caps the upload fan-out at 30 and says so), LinkedIn company Pages 2–20 (images only), Pinterest 2–5 (images only), TikTok up to 35. One url here is simply an ordinary single post. Anything a channel cannot do is REFUSED with the real reason — nothing is ever quietly downgraded to one slide.
linkTitleNothe headline ON the preview card. LINKEDIN NEVER SCRAPES THE PAGE — their Posts API disables URL scraping for API partners outright — so if you do not pass this the card renders UNLABELLED. Fetch the page’s own title and pass it.
visibilityNodefault PUBLIC
captionsSrtNoCLOSED CAPTIONS for a videoUrl post — the SubRip (.srt) CONTENT itself, cue numbers and `00:00:00,000 --> 00:00:02,000` timing lines included, NOT a URL and NOT the plain script (a file with no timings is refused, because LinkedIn would accept it and then silently never show it). Most of LinkedIn is watched with the sound off, so an uncaptioned video is one most of the feed never hears. LinkedIn allows ONE caption file per video and ENGLISH ONLY; it can be attached only WHILE the video is uploaded, never added to a published post; and it is processed asynchronously, so the reply confirms it was UPLOADED and never that it is visible yet. Requires videoUrl — passing it on an image, carousel or link post is refused by name.
allowDuplicateNopost it even though an identical post was just made or attempted. Only pass this when the user genuinely wants the same thing posted twice, or when you have LOOKED at the account and confirmed a timed-out attempt did not land.
idempotencyKeyNoSAFE RETRIES. Publishing can take minutes (a video upload, a carousel of ten slides) and a transport can time out while the post SUCCEEDS — retrying blind is how the same thing gets posted twice. Pass any stable string here and a repeat of the SAME publish returns the ORIGINAL post id instead of posting again (24h). You do not have to: an identical publish is auto-recognised for 10 minutes anyway. If a call times out or errors ambiguously, CALL AGAIN WITH THE SAME KEY — that is the safe move, and it will either report the original post or publish it for the first time. It never posts twice.
organizationIdNonumeric Page id from list_linkedin_pages
linkDescriptionNothe sub-line on the preview card. Same rule as linkTitle: absent means blank, because LinkedIn will not fetch it.
linkThumbnailUrlNoa Hermoso-hosted image used as the card’s picture (uploaded to LinkedIn for you). Without it the card has no image.
videoThumbnailUrlNothe COVER IMAGE for a videoUrl post — a Hermoso-hosted image (a render, or any picture of the user’s via upload_file). Without it LinkedIn adds a system-generated thumbnail, which on an ad is usually whatever the first frame happens to be. Like captions this can only be set WHILE the video is uploaded, never afterwards. Requires videoUrl. This is NOT linkThumbnailUrl, which is the picture on a link-preview card.
Behavior5/5

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

Beyond the sparse annotations (readOnlyHint false, openWorldHint true, idempotentHint false, destructiveHint false), the description discloses crucial behavioral traits: it 'PUBLISHES immediately and PUBLICLY' and mandates user confirmation before calling; it explains that media must be Hermoso-hosted and uploaded to LinkedIn; it details the irreversibility of media attachments (image/video/captions/thumbnail cannot be swapped after publishing) while noting copy can be edited via manage_linkedin_post; and it explains link-preview behavior (LinkedIn never scrapes). These are exactly the kind of behavioral caveats an agent needs and the annotations do not provide. No contradiction with annotations.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Conciseness5/5

Is the description appropriately sized, front-loaded, and free of redundancy?

The description is long but dense, with every sentence earning its place. It front-loads the core purpose and the most critical decision rules (link vs. text, sibling distinction, public nature) before diving into content-type specifics. The structure leads with the action, then the key caveats, then parameter-level detail, and ends with an alert about immutability. No redundant or filler text; it is appropriately verbose for a tool with 18 parameters.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Completeness5/5

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

For a tool with 18 parameters and no output schema, the description covers all critical usage dimensions: media constraints (hosting, mutual exclusion), link-preview requirements, organization selection, idempotency/safe retries, captions and thumbnails, alt text, visibility defaults, duplicate handling, and the distinction from the sibling. It even hints at response behavior (e.g., 'the reply confirms it was UPLOADED and never that it is visible yet' for captions, and idempotent retries return the original post id). Nothing essential is missing for an agent to call it correctly.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Parameters5/5

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

Although schema description coverage is 100%, the description adds substantial semantics beyond each parameter's own text. It explains cross-parameter relationships (linkUrl is mutually exclusive with image/video/carousel; captionsSrt requires videoUrl; videoThumbnailUrl is distinct from linkThumbnailUrl), prescribes how to source parameter values (linkTitle/linkDescription/linkThumbnailUrl must be read from the page; organizationId comes from list_linkedin_pages; media must be Hermoso-hosted via upload_file), and clarifies the meaning of array ordering for carousels. This goes well beyond what the schema descriptions alone provide, giving the agent actionable guidance on how to fill each field correctly.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Purpose5/5

Does the description clearly state what the tool does and how it differs from similar tools?

The description states a specific verb and resource: 'Publish a post to one of the user’s LinkedIn COMPANY PAGES' and enumerates concrete content types (text, image, video, carousel, link post). It explicitly contrasts with the sibling post_to_linkedin ('which publishes to the person’s own profile'), naming the alternative and the selection condition. No ambiguity remains about what this tool does or how it differs from its sibling.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Usage Guidelines5/5

Does the description explain when to use this tool, when not to, or what alternatives exist?

The description gives explicit when-to-use guidance: 'USE linkUrl WHENEVER THE POINT OF THE POST IS A LINK' with the rationale (LinkedIn disables URL scraping), and it warns against using this tool for personal-profile carousels ('ORGANIC CAROUSELS ARE COMPANY-PAGE ONLY — a personal profile cannot publish one and is refused by name'). It also instructs the agent to pick the correct sibling 'and never substitute.' These are clear decision criteria with no reliance on inference.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

Install Server

Other Tools

Latest Blog Posts

MCP directory API

We provide all the information about MCP servers via our MCP API.

curl -X GET 'https://glama.ai/api/mcp/v1/servers/hermoso-ai/hermoso'

If you have feedback or need assistance with the MCP directory API, please join our Discord server