OneUp Today

Throw away the unsent DM draft on a conversation. A conversation that only ever held an unsent first draft is removed with it; otherwise the conversation stays and only its unsent drafts (follow-up included) go, and any queued send is cancelled. Use when the draft is wrong and you do not want to send it.

Discard one specific unsent message by id (a follow-up draft, for example).

Generate a personalized Reddit DM opener for a post or comment using OneUp's DM writer. Creates draft copy only; nothing is sent. Consumes account credits.

Get the objection-handling follow-up draft for a conversation, writing one if there is no current draft and the lead has replied (nothing is written before their first reply; pass regenerate=true for a fresh one). Draft only; nothing is sent. Consumes account credits when it has to write.

Send the user's own reply to a website visitor in a chat that was handed over to them; it appears live in the widget on their site. Confirm the exact text with the user first; once confirmed, make the call.

Write three suggested replies for a conversation, grounded in the original post and the user's product, plus the conversation's objection-handling follow-up draft first when there is one. Nothing is sent; it may also save or refresh that follow-up draft. Consumes account credits.

Get full detail for one agent: its config, the product it promotes, its connected account and its analytics totals (messages are in list_conversations).

Get an agent's audit log (created/edited/paused/resumed/run_now), with actor.

Get one conversation: the lead, its status, the post or comment it started from, and the thread's newest 20 messages (message_count gives the total).

Transcript of one website-chatbot conversation (its newest 40 messages, in reading order), plus the visitor's details.

Whether Front Desk (the chat widget on the user's own website) is set up and live: domain, availability right now, unread conversation count, and knowledge base size. Start here for any Front Desk question.

Get one lead's detail (scores, status, notes, conversation link).

Pipeline summary for the user's leads: totals by status (new, engaged, qualified, trail, paid, disqualified), by source, and by agent. Use this for 'how are my leads doing' before listing anything.

Report whether real Reddit sends are possible right now (send-token freshness, and per agent whether enough posts are gathered for it to run), with plain-English advice on what to do.

One post in full: its whole body, the title and subreddit for Reddit, attached media, the series it belongs to, which agent run produced it, and, once it is live, the URL it was published at with its likes and comments. Use this after list_posts whenever the user asks to read, review, or check on a specific post. Also returns review_url, the page in the app where they can edit and publish it.

Get one reply in full: its text, status, and the post or comment it answers.

Get one run's stats (posts fetched/analyzed/relevant, messages sent).

List replies belonging to ONE agent, optionally filtered by status (draft/posted/discarded/failed). Use list_reply_drafts for the cross-agent view.

List an agent's run history with per-run stats (paginated).

List your agents (Messaging, Email, Reply, Content) with status, platform, schedule and targeting. Each row's agent_label names which kind it is in the user's words: an Email Agent is stored as agent_type 'messaging' on platform 'email', so read agent_label or roster_type, not agent_type, to tell them apart. Rows without agent_type are legacy multi-mode campaigns. paused_note and status_note say in plain words why a quiet agent is quiet (waiting for posts, finished, paused) and what restarts it.

List conversations (unread, draft, or replied) across all agents, or for one agent when campaign_id is given. Use filter=draft to see messages waiting for approval.

The reasons a lead can be disqualified, for use with update_lead_status.

Recent conversations visitors had with the website chatbot, newest first. Filter to only the ones handed over to the user (escalated) or only unread ones.

List leads for you or a campaign, optionally filtered by funnel status.

List your saved message templates (reusable snippets you send often).

List the user's posts and post drafts, newest first: what the Content Agent wrote, plus posts the user drafted in the app. Use it for 'what drafts are waiting on me', 'what did my agent write', or 'what have I published'. Filter by status ('draft' = written but not published, 'published' = live, plus suggested/scheduled/processing/failed), by platform, and by origin ('agent' = written by an agent, 'composer' = the user's own app draft). Returns {count, posts:[{post_id, platform, status, kind, origin, title, subreddit, series, media_count, created_on, scheduled_time, preview}]}. Pass a post_id to get_post for the full text.

Lists the products a Reddit lead search can run for: every product on the user's OneUp Today account, with product_id, name, description and url. start_reddit_lead_search with a product_id finds people on Reddit for one of them. An empty list on a new account means no search has run yet; the first start_reddit_lead_search creates the product.

List public reply drafts your Reply agents have written, across every agent: ones waiting on your approval, plus approved ones still queued to post (queue_state is set on those). Returns reply_id + campaign_id (pass it as agent_id), the draft text, and what it replies to.

List replies that have already been posted, newest first. Optionally narrow to one connected account with social_id. Use this to see what actually went out.

List connected social accounts; Reddit rows include send-token freshness.

Find a conversation by its id alone when you do not know which agent it belongs to. Returns the conversation plus its agent_id, which the other conversation tools need.

Mark one conversation as read, the same as opening it in the app inbox.

Mark one conversation unread again so it resurfaces in the inbox.

Mark one website-chatbot conversation as read, clearing it from the unread badge.

Pause an agent (status -> paused, paused_reason owner_paused): its scheduled runs stop and it stays paused, resumable at any time with resume_campaign. This replaces any automatic pause the agent was in, so only resume_campaign restarts it. Nothing it found is deleted. Idempotent.

Publish a reply draft to Reddit, X, Instagram or Threads from the user's own connected account. It posts publicly right away, except on a Reddit account connected through the OneUp Chrome extension, where the result status is "queued" and it posts when the user's browser is open (tell the user it is queued, never that it was posted; cancel_queued_reply stops it). Confirm the exact text with the user first; once they have confirmed, make the call. Pass reply_text to edit the draft before posting.

Rewrite the unsent draft on a conversation. Pass instruction to steer it (shorter, warmer, answer their question); omit it to regenerate from scratch. Replaces the current draft; nothing is sent. Consumes account credits.

Remove a tracked Reddit text-search term.

Send a reply in an existing DM or email thread from the user's own account; on a Threads conversation the reply is published publicly. It goes to the other person, so confirm the exact text with the user first; once confirmed, make the call. Returns a job id and the send completes in the background, except an X reply, which is queued for the OneUp Chrome extension to send.

Resume a paused agent (status -> scheduled). Idempotent. A messaging agent on a karma-laddered Reddit account stays paused and starts by itself once the ladder completes.