PopMCP
later_instance_Aggregated performance summary across your instances for a date range — one rolled-up total per requested metric.
Later Influence MCP server for your AI
Pull campaign, creator and per-post numbers for any date range, break results down by platform, and see ROI next to earned media value. Reporting only, nothing gets published.
PopMCPLater Influence
Show me engagement by campaign for the spring launch window
Supported Later Influence tools
Every reporting call and watchable threshold: instance, campaign, platform, creator and post performance, time-series and ROI.
later_instance_Time-series of instance performance, bucketed by `granularity` (day/week/month/quarter/year).
later_list_List the reporting instances your API credentials can access.
later_campaign_Per-campaign performance rows for a date range — one row per campaign with the requested metrics.
later_list_Flat list of campaigns across your instances, with basic metadata (id, name, instanceId, description, status live|draft, startDate, firstPostDate, lastPostDate) — NO performance metrics.
later_platform_Performance broken down per platform / content type — one row per contentType (e.g.
later_platform_Return-on-investment / earned-media-value breakdown per content type.
later_creator_Per-creator (influencer) performance rows for a date range.
later_post_Per-post performance rows — the most granular view: one row per published post, with its metadata (id, date, contentType, URLs), the creator who posted it, and the requested metrics.
Connect Later Influence to your AI
Credentials off one Later Influence screen, then an endpoint your client signs into. Nothing to deploy, and nothing to install.
Paste the Client ID and Client secret Later issued you. We swap them server-side for a short-lived token and refresh it automatically, scoped to the reporting instances those credentials can access.
Connection type
your-later-valueClient ID
Paste your Later Reporting API clientIdClient secret
••••••••••••3f7aConnectionVerified
Every call here reads; none write. Nothing to approve, because nothing can be changed: no agent can edit a campaign or create, edit or schedule a post. All nine reporting tools load by default and stay read-only.
Full accessRead-only
Expose full catalog
Approval required for writes
Drop the PopMCP URL into Claude, ChatGPT, Codex or any other supported client. It signs in through PopMCP with OAuth, so your Later Influence credential never leaves the server.
app.popmcp.com/mcp/…
Claude wants to connectApproveAsk for something harmless before you turn writes on. Once it comes back clean, widen the tool set or switch the connection to full access.
Show me engagement by campaign for the spring launch window
later_list_campaignsok
Then just ask for the work
“Show me engagement by campaign for the spring launch window”
- list_campaigns
- campaign_performance
“Rank every creator by engagement for Q3 and flag the bottom ten”
- list_instances
- creator_performance
“Build a month-by-month reach and EMV recap for the board deck”
- instance_performance_over_time
- platform_roi
“Which platform returned the best EMV, and which posts drove it?”
- platform_roi
- platform_performance
- post_performance
Use Later Influence from the assistant you already open
One hosted endpoint, registered once per client. The same tools and the same permissions, whichever window the request comes from.
Claude DesktopCustom connector
Add the PopMCP URL under Settings → Connectors and the tools appear in every chat.
Claude WebCustom connector
The same endpoint, registered in the browser. Nothing to install on the machine you are working from.
ChatGPTCustom connector
Register PopMCP as a custom connector, then let ChatGPT read and update the account.
CodexDesktop & CLI
Wire the endpoint into Codex or the Codex CLI when a task needs live business context, not a fixture.
OpenCodeCoding agent
Register the remote MCP server in OpenCode's config for terminal-first and scripted agent runs.
ZCodeDesktop app
Add PopMCP as an MCP server in Z.ai's desktop app and work the account from the same window.
Explore other MCP servers
Every one of them runs on the same hosted endpoint, with the same controls.
FAQ
Later Influence MCP server questions
What teams ask before putting Later Influence behind an AI client.
Can the agent schedule or publish posts through this connection?
No. This connects the Later Influence Reporting API, which is Later's only public API and is read-only. Later Social, where scheduling lives, has no API at all. The agent reports on posts that already went out; it cannot create, edit or schedule anything.
Where do I get the Client ID and Client secret?
Later issues them for your account once Reporting API access is turned on. If you don't have a pair yet, ask your Later account manager to request Reporting API access for you. The only credentials the connect form asks for are the Client ID and the Client secret.
Do the credentials expire or need re-pasting?
No. PopMCP exchanges your Client ID and Client secret server-side for a short-lived token and refreshes that token automatically, locked to your account. You only revisit the connection if Later rotates or revokes the credential pair.
What data does the connection actually reach?
Rolled-up instance performance for a date range, the same numbers as a time series bucketed by day, week, month, quarter or year, and rows per campaign, per platform and content type, per creator and per published post. There is also a return-on-investment and earned-media-value breakdown by content type. Everything is scoped to the reporting instances your credentials can access, which list_instances will show you.
Does list_campaigns include performance metrics?
No, and this is the thing that trips people up. list_campaigns returns metadata only: id, name, instanceId, description, status (live or draft), startDate, firstPostDate and lastPostDate. For numbers, the agent follows it with campaign_performance, which returns one row per campaign with the metrics you asked for over a date range.
What can PopMCP watch for on its own?
Threshold triggers cover campaign engagement dropping or spiking, program performance falling week over week, a program reach target being passed, a creator underperforming, a single post breaking out, return on investment dropping for a platform, and one platform's share of results shifting. You set the level on each one. PopMCP watches the reporting feed and wakes your agent when a threshold is crossed.
Do I have to run or host the Later Influence MCP server myself?
No. PopMCP hosts it and gives you a Streamable HTTP MCP URL. There is nothing to deploy, no callback URL to register, and no infrastructure to keep running or patch.
Which AI clients can use this MCP server?
Any client that supports a remote Streamable HTTP MCP server — Claude Desktop, Claude Web, ChatGPT custom connectors, Codex and the Codex CLI, OpenCode and Z.ai's ZCode, as well as editors such as Cursor, Windsurf and Zed. Your client signs in to PopMCP with OAuth; the provider credential never leaves the server.
Why do only 9 of the 9 tools load by default?
Loading 9 tool definitions into every conversation burns context before the work starts. PopMCP loads a curated 9-tool set covering the everyday workflows and keeps the rest searchable, so your AI pulls a specialist tool only when a task needs it. You can switch a connection to the full catalog at any time.
What is the difference between a tool and a trigger?
A tool is something your AI calls — there are 9 here. A trigger is something PopMCP watches for you: 8 Later Influence events that can start work on the PopMCP side. Triggers never enter your AI client's context, and none of them ask you to paste a callback URL into Later Influence.
Connect Later Influence to your AI in minutes
Launch a hosted MCP endpoint for Later Influence, keep tool access under control, and start testing real workflow prompts right away.