Pluto Ads

Documentation

Pluto Ads is where your team drafts, reviews and publishes Meta ads, automates the work with loops and hands tasks to the Agent, together with the agents you connect. These guides follow the work from signing in to reading results. The app, your scripts and your agents all use one API, with the same permissions and the same audit log.

Get started

  • Getting started: sign in, the Getting started checklist, finding your way around, inviting teammates, roles and permissions.
  • Connect Meta: what we import when you connect, and bringing live campaigns in with Import from Meta.
  • Connect Pluto Profit: bring your store's profit numbers into Pluto Ads, read-only.
  • Connect Slack: the channels your loops and agents may post in, and what Pluto Ads asks Slack for.

Launch ads

  • Media library: upload images and videos, folders, search and archive.
  • Launches: the board, ad sets and ads, launch settings, bulk edits, tags, statuses, review and sharing with guests.
  • Publishing: review and approval, what gets created in Meta, errors, live changes, pausing, budgets and deleting.
  • Loops: start from a template or build your own automation, choose what the Agent may do, and follow each run and approval.

Measure

  • Analytics: the dashboard, currencies, comparisons, how far back history reaches and launch results.

Work together

  • Collaboration: presence, comments, the inbox, working offline, sync status and conflicts.
  • Guests: invite clients to review and approve the launches you share with them, for free.
  • Agent: ask the in-app Agent to do the work, with your own guidance, skills and approvals.
  • Agents: connect Claude, Cursor or your own agent, what agents can do and how you see their work.

Account

Help and reference

  • Developers: API keys, agent credentials, webhooks and rate limits.
  • Troubleshooting: every error message the app shows, why it happens and what to do.
  • Public API: credentials, idempotency keys, conflicts, batches, launch specs, jobs, webhooks and every route.
  • MCP server: connect Claude, Cursor or your own agent, and the tools it can call.
  • API errors: every error code the API returns, and what to do about it.