Skip to content
Convia ProDocs

For the complete documentation index, see llms.txt. This page is also available as markdown.

Convia for AI agents

Learn the two ways AI agents work with Convia, reading these docs and acting on your workspaces through MCP, plus what an agent can and cannot do.

Last updated
On this page

Convia Pro works with AI agents in two ways: an agent can learn how Convia works by reading these docs, and it can act on your workspaces, reading content and writing drafts, through the Convia MCP server. Nothing an agent writes through MCP is published: posts are drafts until a person schedules them. Only a workspace Owner or Manager, on a plan that includes MCP Connections, can connect an agent.

Two ways an agent works with Convia

Read the docs Act on a workspace
What it is Every docs page as plain markdown, plus an index and a single bulk file A Model Context Protocol (MCP) server at https://mcp.conviapro.com
What the agent gets How features work, settings, limits, how-tos The workspace's own episodes, transcripts, clips, posts, ideas, campaigns and analytics, and the ability to write drafts
Needs an account No Yes: an Owner or Manager, on a plan that includes MCP Connections
Start here Machine-readable docs Connect an AI assistant (MCP)

Most agents use both: the docs to understand what a feature does, and MCP to do the work in a specific workspace.

What an agent can do through MCP

The tools an agent sees depend on the workspace's plan. With them an agent can:

  • Learn the brand. Read the workspace's Brand Brain (voice, audience, content rules, host voice), its prompt templates, its reusable text snippets, its posting schedule and where its posts send people.
  • Find source material. List episodes and guests, read an episode's Episode Prep, read transcripts with speakers and timestamps, search episodes, clips, posts and transcripts, and see which clips have been used least.
  • Check claims. Search transcripts for a figure, or check a list of quotes against an episode in one step, so content rests on what was actually said.
  • See what exists and what is stuck. List posts, the calendar, the Media library, ideas, scripts and campaigns, find open days on a platform's calendar, and list posts waiting on a person, most urgent first.
  • Write drafts. Create posts (with a clip, existing media, an image it adds from a link, or an image or video it uploads), plain draft articles, whole multi-platform campaigns, ideas, scripts and episode records. With Advanced MCP, draft or revise up to 100 posts in one step.
  • Hand drafts to you. Send drafts to your approval queue with a note, marked Submitted by assistant.
  • Revise its own work. Read back and update posts, ideas, scripts, episode details, an episode's prep outline and a campaign plan that has no posts yet. It can fix a small typo without rewriting a post and, with Advanced MCP, fix a word or phrase across a whole campaign.
  • Organize series. On plans that include Series, list and read series, create, rename, archive or restore them, and add items to or remove items from them.
  • Take posts back. Take posts off the calendar or move them to Trash, up to 25 at a time, always with a stated reason.
  • Read results. Post and episode performance, audience growth, the visits posts sent to your website and the searches that bring people to it, on plans with Advanced Analytics.
  • Stage prompt changes. Save a new version of a prompt template for a person to make live.

The full list, with the plan each tool needs, is in the MCP tool reference.

What an agent cannot do

Convia keeps decisions that publish, approve or declare something with a person.

  • Nothing is published through MCP. Every post an agent creates is a draft. A draft with a date sits on the calendar until a person schedules it, and a campaign stays a draft until a person activates it. An agent can send drafts to the approval queue, but cannot approve or schedule them. Editing a post a person already scheduled leaves it scheduled.
  • Published and approved posts are off limits. An agent cannot edit, unschedule or trash them. An edit to a post waiting for approval sends it back to draft, so the change is reviewed. The one exception: with Advanced MCP, an agent can change the title, description, chapters or tags of a video Convia uploaded to YouTube, after showing you the new copy and getting your go-ahead.
  • Some publish settings stay with you. An agent can set an Instagram post's format, whether a reel shows in the feed, and trial reels, plus YouTube category, playlists, tags and audio language, and a Pin's board. It cannot set TikTok privacy or branded-content disclosures, or YouTube's Made for Kids, privacy status or publish time.
  • Images and videos only. An agent can add PNG, JPEG, WebP or GIF images up to 10 MB to the Media library, and MP4 or QuickTime videos up to your plan's file limit (at most 1 GB) from a client that can upload files, such as Claude Code. It cannot add documents or audio.
  • Live AI prompts are protected. An agent cannot change the prompt Convia uses now. It can only stage an alternative or a draft for a person to make live.
  • No deletions beyond posts. Ideas are archived rather than deleted, and there are no delete tools for scripts, episodes or campaigns.
  • Mostly read-only settings. Brand Brain, text snippets and posting schedule settings can be read, not changed. An Owner's or Manager's agent can add or change destinations.
  • Some work stays in the app. Campaign goal results, site health audits, the weekly website growth review, recording campaign spend and choosing which Google Analytics key events count as conversions are done in Convia, not through MCP.
  • Only the workspaces you choose. A connection reaches only the workspaces ticked when it was approved, or the workspace a token was created in. With several, the agent names one on every request and is refused rather than allowed to guess.

Who can connect an AI agent through MCP

  • Role. Only a workspace Owner or Manager can create a token or connect a client, and the role is checked on every request.
  • Plan. MCP Connections comes with Grow and above, and Advanced MCP with Scale and above. Some tools also need a product feature on the plan, such as Advanced Analytics or Magic Campaign. See Plans & billing.
  • Clients. Claude on the web, iPhone and Android and Claude Desktop connect with a custom connector that signs in to Convia and can reach several workspaces. Claude Code and a Claude Desktop config file use a workspace token.

Details are in MCP access, permissions & audit.

MCP credits, limits and accountability

Using MCP spends no AI credits. The agent writes the copy on your own subscription with that AI assistant, and Convia stores the result.

Each workspace has a limit on how much connected assistants can do: 300 requests an hour and 100 changes a day, or 1,000 and 500 with Advanced MCP. Past it, the assistant is asked to wait and try again.

Every tool call is recorded on the Activity tab of Integrations → MCP Connections (under Workspace in the launcher), with the tool, the result and the time it took. Tokens can be revoked and connected apps disconnected at any time, and removing a member or changing their role stops their access on the next request.

Where to start with AI agents

  1. Read Connect an AI assistant (MCP) to connect Claude or another client.
  2. Have the agent call get_started first. It returns this page, the workspaces the connection reaches, how each one is set up and which tools fit common jobs on your plan, and suggests what to do first. See Start a session with get_started.
  3. Try the recipes in Agent workflows, starting with drafting one post.
  4. Keep the MCP tool reference and the Glossary at hand for exact names and terms.
  5. Point your agent at Machine-readable docs so it can look up how any feature works.