com.makehybrid/mcp

The tasks in one project: with no since, the ones that are not done; with since (milliseconds), the ones Jira changed after it, oldest first. Each carries task_id, title, notes, status (open, doing or done), assignee, due and updated_at. -- MakeHybrid API plan, GET /jira/v1/connections/{connection_id}/projects/{project_id}/tasks, group jira.

The chats this account has heard from, newest first, with chat_id and whether each is a person, a group or a LINE room. LINE lists none itself, so a chat is known only once it has written. -- MakeHybrid API plan, GET /line/v1/connections/{connection_id}/chats, group line.

What LINE says this account is right now: its display name, its basic id and whether its chat is still on. -- MakeHybrid API plan, GET /line/v1/connections/{connection_id}, group line.

Where a member pastes the channel access token and channel secret of their LINE Official Account: this answers with that page and nothing else, because a secret never goes through this API. The connection appears here once they have finished. -- MakeHybrid API plan, POST /line/v1/connections/link, group line.

The LINE Official Accounts this organization connected, with the connection_id every other call names and the account name LINE shows. -- MakeHybrid API plan, GET /line/v1/connections, group line.

List Managed Company Pages -- MakeHybrid API plan, GET /linkedin/v2/{account_id}/linkedin/company/pages, group linkedin.

List Available Contracts -- MakeHybrid API plan, GET /linkedin/v2/{account_id}/linkedin/contracts, group linkedin.

Get InMail Credits -- MakeHybrid API plan, GET /linkedin/v2/{account_id}/linkedin/inmail-credits, group linkedin.

List User Job Postings -- MakeHybrid API plan, GET /linkedin/v2/{account_id}/linkedin/jobs, group linkedin.

List all Recruiter Hiring Projects -- MakeHybrid API plan, GET /linkedin/v2/{account_id}/linkedin/projects, group linkedin.

List Job Postings -- MakeHybrid API plan, GET /linkedin/v2/{account_id}/linkedin/recruiter/jobs, group linkedin.

Perform Classic Search from URL -- MakeHybrid API plan, POST /linkedin/v2/{account_id}/linkedin/search, group linkedin.

List Search Parameters -- MakeHybrid API plan, GET /linkedin/v2/{account_id}/linkedin/search/parameters, group linkedin.

Perform People Search -- MakeHybrid API plan, POST /linkedin/v2/{account_id}/linkedin/search/people, group linkedin.

Perform Posts Search -- MakeHybrid API plan, POST /linkedin/v2/{account_id}/linkedin/search/posts, group linkedin.

Read a LinkedIn company page you administer, as one of your connected accounts (query account_id): its name, tagline, overview (the About text), website, page address, industries, company size, company type, phone, year founded, specialties, custom button, logo and cover, and the sizes, types and buttons this page accepts. company_id is the page's number or its urn. -- MakeHybrid API plan, GET /linkedin/pages/{company_id}, group pages.

Replace a LinkedIn company page's logo, as one of your connected accounts (account_id): send image_url, a public https address of a PNG or JPEG, or upload the file as multipart field image. A square picture of 400 by 400 or more works best. -- MakeHybrid API plan, POST /linkedin/pages/{company_id}/logo, group pages.

List all Post's Comments -- MakeHybrid API plan, GET /linkedin/v2/{account_id}/posts/{post_id}/comments, group posts.

Create a Post -- MakeHybrid API plan, POST /linkedin/v2/{account_id}/posts, group posts. THE WORDS ARE YOURS TO WRITE. This API carries them to the page and keeps them; it holds none of the member's pricing, product, launch dates or announcements, so missing a fact, draft anyway and ask for that one line alongside the draft. HOW A POST FOR THIS MEMBER MUST READ: First line under eight words and no emoji in it. No emoji anywhere unless the member has asked for them -- no rocket, no tick marks, none as bullets. No markdown: LinkedIn renders none of it, so **bold** arrives on the page as asterisks. NEVER a long dash (the em dash). Write two sentences, or use a comma. Contractions, sentences of uneven length, and one concrete thing -- a number, a name, a week. No Ready to...? close, no Thoughts?, no call to action unless the member asked for one. Stop on the last real sentence. Show it to the member before anything is published.

Get a Post -- MakeHybrid API plan, GET /linkedin/v2/{account_id}/posts/{post_id}, group posts.

List all Post's Reactions -- MakeHybrid API plan, GET /linkedin/v2/{account_id}/posts/{post_id}/reactions, group posts.

Where a customer conversation nobody linked arrives: room_id (null means one is chosen on the next message, and a room named Customer conversations is made if there is none) and off. -- MakeHybrid API plan, GET /rooms/arrivals, group rooms.

The platforms a room's conversation can be carried to (linkedin, discord, slack, clickup, whatsapp, chatwoot, line, x, webhook, inbox, webchat) and the organization's accounts on each. -- MakeHybrid API plan, GET /rooms/channels, group rooms.

Make a project room: name (up to 80 characters), guests_can_ask (whether invited guests may ask its AI agents). It starts with one chat. -- MakeHybrid API plan, POST /rooms, group rooms.

The organization's AI model keys for rooms (id, provider, label, last4). A new key is added on the rooms page, never through the chat. -- MakeHybrid API plan, GET /rooms/keys, group rooms.

List the organization's project rooms, each with url (where a person who is in it opens it), its chats (id, name) and shared_url when it is on the web. -- MakeHybrid API plan, GET /rooms, group rooms.

A room's messages, oldest first: chat (a chat id; the first chat when absent), after (a seq, for only what is new), limit up to 200. -- MakeHybrid API plan, GET /rooms/{id}/messages, group rooms.

Say something in a room as the organization's owner: text, chat_id (the first chat when absent). Its agents may answer, and its linked channels carry it out. Ask before writing where people read. -- MakeHybrid API plan, POST /rooms/{id}/messages, group rooms.

The MCP servers a room's AI agent can be given: the organization's own agents, Top Voice when connected, and servers added by address. -- MakeHybrid API plan, GET /rooms/servers, group rooms.

Add a task to a room: title, notes, status, ready (whether it is specified well enough for an assistant to finish on its own), assignee, due (YYYY-MM-DD), priority, links [{title, url}]. -- MakeHybrid API plan, POST /rooms/{id}/tasks, group rooms.

A room's tasks: title, notes, status (open, doing, done), ready, assignee, due, priority (low, normal, high, urgent), links, and the channel that made one. ready=true gives only the ones waiting to be taken. -- MakeHybrid API plan, GET /rooms/{id}/tasks, group rooms.

The work waiting to be taken, across every room: each task with its room. A room is a conversation around making and closing tasks, and a ready one is specified well enough to finish without asking - what it is, what done looks like, and the links it needs. This is how a person carries work out of a room and into whichever assistant they work in: take one, do it there, then PATCH its status to done. -- MakeHybrid API plan, GET /rooms/work/ready, group rooms.

What this organization still has to do before its LinkedIn is actually working for it, and the one call that does the next bit. Answers ready, next and steps: connect a LinkedIn account; add the organization's own AI key (which goes in on a page, never through you); start a content campaign so the account posts regularly; start a boost or Top Voices campaign so the right people see it; open a project room and point the campaigns at it; connect the channels the answers come back on; set the campaigns running. Each step says whether it is done, why it matters, and the exact call to make, worked out from what this organization actually has. A step that reads unknown is one this deployment cannot see either way - never treat it as done. Call this at the start of a conversation about setting anything up, and again after each step. -- MakeHybrid API plan, GET /linkedin/setup, group campaigns.

A link to Slack's own page that installs the app in a workspace: give it to someone who may add apps there. Once they allow it, the workspace is the organization's for this API and for project rooms. -- MakeHybrid API plan, POST /slack/v1/connection/link, group slack.

The Slack workspaces the organization installed the app in, with their team_id. -- MakeHybrid API plan, GET /slack/v1/workspaces, group slack.

A workspace's channels: every public one, and the private ones the app was invited to, with channel_id and whether the app is a member. limit up to 1,000 and cursor to page. -- MakeHybrid API plan, GET /slack/v1/workspaces/{team_id}/channels, group slack.

A person in the workspace: their display name, real name, time zone and whether they are a bot. Message user_id values are these. -- MakeHybrid API plan, GET /slack/v1/workspaces/{team_id}/users/{user_id}, group slack.

Find the LinkedIn top voices on a topic, as one of the organization's connected accounts. Send query.account_id and query.topic, and query.limit up to 50. Each one comes back with a score out of 100 and the parts it is made of - reach (how many follow them), resonance (what a post of theirs earns), conversation (how much of that is comments rather than taps) and cadence (how regularly they publish) - along with the figures behind them and whether this organization already watches them. LinkedIn's own Top Voice badge is reported as badge when the profile exposes it; badge null means it did not say, never that there is none. -- MakeHybrid API plan, GET /linkedin/topvoices, group campaigns.

Get a User Profile -- MakeHybrid API plan, GET /linkedin/v2/{account_id}/users/{user_id}, group users.

Get a User Chat -- MakeHybrid API plan, GET /linkedin/v2/{account_id}/users/{user_id}/chat, group users.

List all pending Relation / Follow requests -- MakeHybrid API plan, GET /linkedin/v2/{account_id}/users/me/relation-requests, group users.

Report a Profile Visit -- MakeHybrid API plan, POST /linkedin/v2/{account_id}/users/visit-profile, group users.

The organization's AI agents a website chat or a conversation can be given to, each with the room it belongs to. They answer visitors without tools, on the organization's own AI key. -- MakeHybrid API plan, GET /webchat/v1/agents, group webchat.

The organization's rooms, for room_id: a website chat's conversations become tasks there, one per visitor, assigned as the conversation is. -- MakeHybrid API plan, GET /webchat/v1/rooms, group webchat.

Add a website chat: name (what visitors see), color (#rrggbb), greeting, sites (host names that may show it, *.example.com for subdomains; none means any), ask_email (off, optional, required), room_id and agent_id. The answer's snippet goes before </body> on every page. -- MakeHybrid API plan, POST /webchat/v1/widgets, group webchat.

The organization's website chats, each with its id, look, sites, the room its conversations become tasks in (room_id), the AI agent that answers first (agent_id) and the snippet to paste into a website's pages. -- MakeHybrid API plan, GET /webchat/v1/widgets, group webchat.

List this organization's webhook endpoints: each one's id, the URL events are sent to, which events it asked for, and whether it is still delivering. Signing secrets are never returned here -- a secret is shown once, when the endpoint is created. -- MakeHybrid API plan, GET /linkedin/webhooks/endpoints, group webhooks.

Open the AI inbox: a webhook kept for your AI instead of sent anywhere, receiving every event: your connected LinkedIn accounts, your shared inbox and your X account. Opened for you when an AI connection is created; calling it again answers the same inbox. -- MakeHybrid API plan, POST /linkedin/webhooks/inbox, group webhooks.

Read the AI inbox: what happened on your connected LinkedIn accounts (messages received and sent, new relations, account status changes and checkpoints), in your shared inbox (inbox.* events) and on your X account (x.message_*), oldest first, kept 30 days. Pass query.after = the next value from your last read to see only what is new, query.event to filter, query.since for a time. -- MakeHybrid API plan, GET /linkedin/webhooks/inbox, group webhooks.

Whether events are flowing at all, before you debug one endpoint: receiving_events says whether this server is subscribed upstream, upstream_configured whether the connection behind every event exists, and endpoints and accounts_connected how many of each this organization has. receiving_events false with endpoints above zero means nothing will ever arrive, however correct the endpoint is. -- MakeHybrid API plan, GET /linkedin/webhooks/status, group webhooks.