Skip to main content
Glama

Eurekuh

Create, schedule or publish a post

create_post
Destructive

Create a post for one or more connected networks.

mode "draft" saves it and publishes nothing. mode "schedule" queues it for scheduledAt. mode "now" PUBLISHES IMMEDIATELY to live public accounts. That cannot be undone from here — the post has to be deleted on the network itself. Prefer draft or schedule unless the user has asked for it to go out now.

Posting to X immediately is refused by default because an X post is a billable charge; schedule it instead. Run check_limits first if you are unsure the caption fits.

Tool call template

Replace each placeholder with a value matching the input schema. Only required parameters are included; add optional parameters as needed.

{
  "id": 1,
  "jsonrpc": "2.0",
  "method": "tools/call",
  "params": {
    "arguments": {
      "mode": "<mode>"
    },
    "name": "create_post"
  }
}

Input Schema

NameRequiredDescriptionDefault
modeYes
mediaNomediaId values from upload_media or upload_file. The first image is the cover.
captionNoThe caption. A line containing only "---" splits it into a thread on the networks that support one.
igTrialNoInstagram video reels only: post as a trial reel, shown to non-followers first. "SS_PERFORMANCE" shares it to followers if it performs well; "MANUAL" leaves sharing to the user in the Instagram app. Not for stories or photos.
topicTagNoThreads only: one topic tag, e.g. "Travel Threads". 1–50 characters, no "." or "&". Other networks ignore it.
igDmReplyNoWhat igDmKeyword sends back.
placementNoInstagram and Facebook only: "story" posts the photo or video as a Story (24 hours, no caption shown) instead of to the feed. Default "feed". TikTok and the other networks have no stories API.
platformsNoNetworks to post to, using each one's default account.
thumbnailNoYouTube video posts only: a JPG or PNG mediaId from upload_file or upload_media, set as the custom thumbnail. Applied once the video is live and YouTube accepts it (often only after processing), retried for about 7 hours. A 9:16 image suits a Short. The channel must be verified for custom thumbnails; get_post_status shows the result.
accountIdsNoSpecific account ids from list_accounts. Use this when the user has more than one account on a network.
visibilityNoWho may see the post. Default "public". TikTok: public, followers, friends, private. YouTube: public, unlisted, private. LinkedIn: public, connections. Every other network only posts publicly, so anything but public is refused there.
igDmKeywordNoInstagram, Facebook and Threads only. Anyone who comments this word on the post gets igDmReply as a private message, once each, within 7 days; on Threads, which has no DMs, as a public reply. Needs igDmReply too.
scheduledAtNoISO 8601, in the future. Required when mode is "schedule".
firstCommentNoPosted as the first comment (a reply, on Threads and Bluesky) where supported. When: see firstCommentTrigger.
dmPublicReplyNoOptional. Posted publicly under the comment once the DM is sent, e.g. "Sent you a DM 📩". Leave out for no public reply.
igDmReplyAltsNoOptional other versions of igDmReply (up to 9). Each commenter gets one of igDmReply and these at random, so a busy post doesn't send the identical DM every time.
organizationIdNoWhich organization to act in: an id from list_organizations. Leave out for this connection's default organization.
pinterestBoardNoPinterest only: the board to pin to, by name or id. Left out, or no board matches, it goes to the account's first board.
firstCommentTriggerNoWhen firstComment goes out. "now" (default): right after the post is live. "after": firstCommentThreshold minutes after it is live (e.g. 240 = 4 hours; at most 7 days). "views" / "likes": once the post has firstCommentThreshold of them — threads and instagram only, read every 5 minutes at first and less often as it ages; if it never gets there in 7 days the comment is not posted and get_post_status says why, unless firstCommentFallbackMinutes is set. A comment never fails the post.
firstCommentThresholdNoMinutes for "after", the count for "views" / "likes". Required unless the trigger is "now".
firstCommentFallbackMinutesNoOnly with "views" / "likes": whichever comes first. The comment goes out once the post has firstCommentThreshold of them, or this many minutes after it is live, even if the network gives no reading (e.g. threshold 5 + 300 = 5 likes or 5 hours). A whole number, 1 to 10080 (7 days). null removes it.

Schema Changelog

Changes observed during successful MCP inspections.

  1. First observed

TDQS

A4.8/5.0
Behavior5/5

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

Annotations already flag destructive, openWorld, and non-idempotent, but the description adds irreplaceable context beyond them: "now" publishes to live public accounts and "cannot be undone from here — the post has to be deleted on the network itself." It also discloses that immediate X posting is refused by default because it is a billable charge, which no annotation conveys.

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?

Front-loaded with purpose, then a compact per-mode block, then the danger warning, then the X refusal caveat. Every sentence carries distinct operational information; there is no filler and the destructive case is emphasized rather than buried.

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

Completeness4/5

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

For a 21-parameter, high-stakes publish tool with no output schema and 95% schema coverage, the description covers the decisive risks: mode semantics, irreversibility, the X billing refusal, and the caption-length check. It does not mention what is returned (e.g. an id usable with get_post_status) or which networks require media, which would help close the loop after a call.

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

Parameters4/5

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

Schema coverage is 95%, so the baseline is 3, but the description adds real meaning to the one required parameter by explaining what each mode value actually does — the schema only lists the enum values without per-value semantics. The pointer to check_limits for caption length also adds cross-tool meaning, though the other 20 parameters are left entirely to the schema.

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 first sentence states a specific verb and resource ("Create a post") plus the scope ("one or more connected networks"), and the three mode labels make the operation's range immediately clear. It is easily distinguished from siblings like update_post or set_first_comment, which operate on already-created posts.

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?

Each mode gets an explicit when-to-use statement, plus a default recommendation ("Prefer draft or schedule unless the user has asked for it to go out now") and a hard precondition ("Run check_limits first if you are unsure the caption fits"). It also names a case where the tool refuses (immediate X posting) and directs the agent to schedule instead, so alternatives and exclusions are covered.

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

Try in Browser

Glama MCP Gateway

Add one secure layer between your agents and this server.

Resources