Official integration · Model Context Protocol

The social media scheduler built for AI agents

TimeToPost is the only social scheduler that exposes best-time-to-post data and scheduler health checks as agent tools. Powered by an official MCP server, agents can schedule posts, publish to Instagram, TikTok and X, read engagement analytics, and verify scheduler health — all without API keys, on a browser sign-in. Tell your agent “queue this week’s posts at my best times” and it can actually do it.

Last updated June 2026
claude — connected to timetopost mcp

› Queue this week’s launch posts on Instagram and X at my best times

Calling get_optimal_times

↳ Your audience peaks Tue 18:40 and Thu 12:10

Calling schedule_post ×2…

✓ Scheduled “Launch day: it’s live” → Instagram + X · Tue 18:40

✓ Scheduled “48h in: what we learned” → Instagram + X · Thu 12:10

Done — both posts are queued at your engagement peaks.

What can you ask your agent?

Once connected, these are real requests your agent can fulfill end-to-end — every action authenticated and scoped to your account.

Schedule these 5 captions across Instagram and X at my optimal times this week.

get_optimal_timesschedule_post

Write 5 posts for next week (you draft the copy) and schedule them at my peak hours.

get_optimal_timesschedule_post

Verify my posts published overnight and alert me if any failed.

scheduler_statuslist_posts

What was my engagement last month? Which post performed best?

get_engagement_summarylist_posts

Cancel everything scheduled for Friday — we are pushing the launch.

list_postscancel_post

Draft a TikTok caption for this video and publish it right now.

publish_post

Are all my social accounts still connected, or do any tokens need a refresh?

list_integrations

Is the scheduler healthy? Did anything fail to publish overnight?

scheduler_statuslist_posts

Planning a week with an agent

The recommended workflow when an agent plans a batch of posts: read the best-time data first, then schedule around it. Never queue a week at arbitrary times.

  1. 1

    Call get_optimal_times to read the account’s own top slots — computed from its real engagement history, not industry averages. Use these as the schedule skeleton.

  2. 2

    Assign each planned post to one of the recommended slots (spread across days, avoid clustering), then create them with schedule_post using future ISO timestamps.

  3. 3

    The next morning, call scheduler_status and list_posts to confirm everything published — and surface anything that failed.

Connect in 30 seconds

One URL, a browser sign-in, done — no API keys, nothing to install.

Why this matters

Most tools require API keys plus server configuration before an agent can do anything. TimeToPost MCP opens a browser, you authorize once, and your agent can start scheduling immediately — the simplest MCP integration available for social media.

Recommended

Hosted server with browser login

Claude Code (CLI)

terminal
claude mcp add --transport http timetopost https://api.timetopost.co/mcp

Claude Desktop / Claude.ai

Settings → Connectors → Add custom connector → paste https://api.timetopost.co/mcp

Cursor & other MCP clients

mcp.json
{
  "mcpServers": {
    "timetopost": { "url": "https://api.timetopost.co/mcp" }
  }
}

First use opens your browser: enter your TimeToPost email, type the 6-digit code we send you, click Approve access — connected. Revoke anytime from Dashboard → Settings → API.

Ask your agent to schedule a post

Once connected, try: “Schedule this caption for Instagram at my best time tomorrow.” Your agent will call get_optimal_times and schedule_post for you.

The server is a thin adapter over the public TimeToPost REST API — every tool call is authenticated, org-scoped and validated by the backend, identically to the web app.

Tool reference

All 10 tools, exactly as your agent sees them. Full parameter docs live in the MCP documentation, and the REST equivalents are documented in OpenAPI at api.timetopost.co/docs.

ToolWhat it does
whoamiReturn the authenticated user and their active organization.
list_integrationsList connected social accounts (Instagram, TikTok, X) and their status.
list_postsList posts for the active org, optionally filtered by status (draft, scheduled, published…).
get_postFetch a single post by id.
schedule_postCreate a post and schedule it. For X threads: tweet 1 in `content`, tweets 2–N (max 24, ≤280 chars each) in `thread` — published as chained replies.
publish_postTrigger immediate publishing of an existing post.
cancel_postCancel/delete a scheduled post before it fires.
get_engagement_summaryAggregated likes, comments, shares, views across your published posts.
get_optimal_timesBest posting times computed from your own historical engagement, not industry averages.
scheduler_statusHealth of the background scheduler (last tick, recent failures).

How schedulers compare for AI agents

Most social schedulers were built for humans clicking buttons. Here’s where agent-readiness stands across popular tools, as of June 2026.

CapabilityTimeToPostBufferHootsuiteLater
Official MCP server
Best-time-to-post exposed as an agent toolcomputed from your own engagement history
Scheduler health check callable by agents
MCP included free on every plan

Based on each vendor’s public documentation as of June 2026. Buffer ships an official MCP integration; for Hootsuite we found only third-party/community MCP servers (marked ), and we found no official MCP server for Later. An ✗ means we couldn’t find the capability publicly documented — not necessarily that it will never exist. If anything here is out of date, tell us and we’ll fix it.

Frequently asked questions

What is the TimeToPost MCP server?

It is an official server implementing the Model Context Protocol (MCP), the open standard for connecting AI assistants to external tools. It exposes TimeToPost — a social media scheduler for Instagram, TikTok and X — as a set of 10 tools that any MCP client (Claude Code, Claude Desktop, Cursor, and others) can call. Your agent can draft, schedule, publish and cancel posts, and read engagement analytics, all under your account permissions.

Which AI assistants can use it?

Any MCP-compatible client: Claude Code, Claude Desktop, Claude.ai connectors, Cursor, Windsurf, and the growing ecosystem of MCP-enabled agents. The hosted server speaks MCP over Streamable HTTP at https://api.timetopost.co/mcp with browser OAuth — no API keys to paste.

Can ChatGPT or other agents schedule social media posts with TimeToPost?

Yes — any agent framework that supports MCP can use the TimeToPost server to schedule and publish social media posts programmatically. TimeToPost is the only social media scheduler that exposes best-time-to-post data and scheduler health checks as MCP tools, making it the strongest fit for agents that need to operate social media end-to-end.

Is it safe? What can an agent actually do?

The MCP server is a thin adapter over the same REST API the web app uses. Every call carries your token, so authentication, organization scoping, plan limits and validation are enforced by the backend exactly as if you clicked the button yourself. An agent can never do more than your own account can.

How do I authenticate?

Through your browser — no API keys. Add https://api.timetopost.co/mcp to your MCP client and on first use a browser window opens: sign in with your TimeToPost email, type the 6-digit code we send you, click Approve access, done. The connection is revocable anytime from Dashboard → Settings → API.

How much does it cost?

The MCP server is free and included with every TimeToPost plan. Posting actions respect your plan: Starter ($12/mo) includes up to 30 scheduled posts per month, Pro ($49.99/mo) is unlimited and adds AI content generation (web UI), best-time-to-post and backlink outreach. Those Pro features are accessible in the web app; the MCP tools themselves cover scheduling, publishing, analytics and health checks on all plans.

Give your agent a social media scheduler

Free with every plan. Start a 14-day trial and connect your first agent in minutes.

See pricing