Give your AI assistant the power to publish
Your AI plans and writes. SkedCast checks, schedules, paces and publishes — and proves every post went live. Connect Claude or any MCP client to 58 tools, secured by OAuth.
Writing the post is the easy part
Publishing it reliably, to many accounts, inside every platform’s rules is the hard part. That is the part SkedCast does.
Your AI assistant
- Decides what to say, and to whom
- Writes and revises the posts
- Chooses when, using best times and limits
- Reads results and plans the next round
SkedCast
- Checks every post against each platform’s rules
- Schedules it on every destination at once
- Paces each account inside platform limits
- Publishes, retries, and stores proof it went live
What your assistant can do
31 tools that read and 27 that act — the same workspace your team uses.
Plan and publish
Draft a post, check it against every destination’s rules, then schedule it or save a draft — one post fanned out to many accounts.
- validate_post
- create_post
- update_post
- reschedule_post
- cancel_post
Run the calendar
Answer “what goes out this week”, move a whole campaign by a day, or retry what failed.
- list_scheduled_posts
- list_post_targets
- shift_campaign
- retry_post
Pick the right time
Read each account’s posting limits and best times before choosing a slot.
- get_best_times
- get_posting_limits
- get_capabilities
Learn from results
Pull workspace, client, account and post analytics, and reports your team already set up.
- get_analytics_overview
- get_client_analytics
- get_post_analytics
- list_reports
Work with media
Find assets in your library or bring new ones in from a link.
- list_library_assets
- import_media
- get_media_asset
Keep humans in charge
See the approval queue and record a decision — only when the connected person is allowed to.
- list_approvals
- get_approval
- decide_approval
Safe to hand to an assistant
Every rule below is enforced by the server, not left to the model’s good behaviour.
You approve the connection
OAuth 2.1 with PKCE. You sign in, see what the assistant is asking for, and can disconnect it any time.
It acts as you — no more
Your role, your plan limits and your client access apply to every call. An agent can’t reach a client you can’t.
Write tools ask first
27 tools change something, and each is marked so the assistant confirms before calling it.
Approvals still apply
A post an agent creates goes through the same approval workflow as one from your team.
Connect in a minute
No SDK, no code. Add SkedCast as a remote MCP server in your assistant and sign in.
Add the server
In your MCP client, add a remote server with the URL
https://api.skedcast.com/mcp.Sign in with SkedCast
Your assistant opens the SkedCast sign-in. Review the permissions it asks for and approve.
Ask for what you need
“What’s scheduled for Acme this week?” or “Move the launch campaign to Thursday.”
Built to be found by agents
SkedCast publishes the documents agents use to discover a service and how to sign in to it. Open any of them — they are public.
- MCP server cardThe MCP endpoint, transport and tool list, for MCP clients.skedcast.com/.well-known/mcp/server-card.json
- Agent cardAn A2A agent card: what SkedCast can do, derived from the same tool list.skedcast.com/.well-known/agent-card.json
- OAuth protected-resource metadataWhere an agent finds the authorization server (RFC 9728).api.skedcast.com/.well-known/oauth-protected-resource/mcp
- llms.txtA plain-text map of the product and developer docs for language models.skedcast.com/developers/llms.txt
AI agents FAQ
- Which AI assistants can connect to SkedCast?
- Any client that speaks the Model Context Protocol over Streamable HTTP with OAuth — including Claude and a growing list of agent frameworks and IDEs. You add https://api.skedcast.com/mcp as a remote MCP server and sign in once.
- Can an AI agent post without my permission?
- No. You approve the connection with OAuth, seeing the permissions it asks for, and the agent acts as you — with your role, your plan limits and your workspace’s approval rules. Tools that change something are marked as write tools so the assistant asks before calling them, and a post an agent creates goes through the same approval workflow as one from your team.
- What can the agent actually do?
- The server exposes 58 tools: 31 that read (posts, accounts, analytics, best times, approvals, reports) and 27 that act (create and reschedule posts, move campaigns, import media, decide approvals, manage webhooks). The full list, with the scope each one needs, is in the developer docs.
- Which plan do I need?
- Connecting an AI assistant is included from the Solo plan. Unattended automation with API keys — from a script, a server or CI — is included from Studio.
- Why use SkedCast instead of letting the AI post directly?
- Writing the post is the easy part. Publishing it reliably is not: every platform has its own format rules, rate limits, token refreshes and failure modes, and posting to many accounts safely needs pacing. SkedCast handles validation, scheduling, pacing, retries and proof that a post really went live, so the agent can focus on what to say.
Let your assistant run the publishing
Connect it once. It plans and writes; SkedCast gets every post out, on time and inside the rules.