CreateScheduledAgent
ChatGPTtasks/create - Create a Task. Requires editor access and the configuration:write permission.
Overview
Access CloudX publisher performance directly from ChatGPT to track revenue, fill rate, and bidder contributions across your apps. Pull dashboard summaries, drill into ad units, compare demand partners, and export raw data — all in natural language. Filter by time range, country, or device OS to investigate issues without leaving the conversation.
Tools
DeleteScheduledAgent
ChatGPTtasks/delete - Delete a Task. Requires editor access and the configuration:write permission.
GetAuctionBids
ChatGPTauction/bids - List bid and non-bid rows matching the requested filters. Requires the auctions:read permission.
GetAuctionList
ChatGPTauction/list - List recent auctions matching the requested filters. Requires the auctions:read permission.
GetAuctionRounds
ChatGPTauction/rounds - Inspect rounds for an auction or aggregate round metrics. Requires the auctions:read permission.
GetAuctionShow
ChatGPTauction/show - Inspect a single auction and optional attached details. Requires the auctions:read permission.
GetConfigHistory
ChatGPTconfig/history - List up to 200 of the most recent config history rows with publish-time diff metadata. Results are newest first and are not an exhaustive account history. Requires the configuration:read permission.
GetConfigShow
ChatGPTconfig/show - Fetch a published config version, draft, or the live config. Requires the configuration:read permission.
GetConfigValidate
ChatGPTconfig/validate - Validate a live, published, or draft configuration using server-side account context. Requires the configuration:read permission.
GetManagedPricingAvailability
ChatGPTList dynamic pricing availability - List ad-unit pricing modes and eligibility reasons. Requires configuration:read.
GetReportABTest
ChatGPTreport/ab-test - Fetch A/B test metrics split by control and test variants. Requires the reports:read permission.
GetReportAdUnits
ChatGPTFetch per-ad-unit performance for the requested time range.
GetReportApps
ChatGPTFetch per-app performance for the requested time range.
GetReportBidders
ChatGPTFetch per-bidder performance for the requested time range.
GetReportBreakdown
ChatGPTreport/breakdown - Fetch grouped reporting rows for the requested time range. Requires the reports:read permission.
GetReportDashboard
ChatGPTFetch dashboard summary and chart data for the requested time range.
GetReportExport
ChatGPTFetch CSV-schema export rows for the requested time range.
GetReportTotalRevenue
ChatGPTRead the Total Publisher Revenue report shown on the Total Revenue page for onboarded apps. Prefer it for available app revenue, including unqualified requests and daily reports. Also use it for total publisher revenue, mediation revenue, CloudX revenue, and share of wallet. Revenue and impression coverage is broader than other measures; do not treat missing mediator measures as zero. Label revenue as finalized Total Publisher Revenue and state dataCompleteThrough. Report metadata lists supported measures and dimensions. The response reportLink is the canonical live chart link; copy it verbatim and never construct or modify it.
GetScheduledAgent
ChatGPTtasks/get - Get one Task. Requires the configuration:read permission.
ListScheduledAgents
ChatGPTtasks/list - List the account's Tasks. Requires the configuration:read permission.
PostConfigEdit
ChatGPTconfig/edit - Apply typed app, ad unit, ad unit group, network mapping, list, tag, test-device, line-item, and A/B test config edits. Requires the configuration:write permission. Ordinary config edit actions apply to a draft: if config_id points to a draft, that draft is edited in place; otherwise a new draft is created. A/B test create also creates a draft. A/B test update, start, end, promote, and delete actions publish the resulting configuration as live and must be submitted as the only operation in the request. Ad-unit and line-item bidfloor values are decimal USD CPM; for example, 1.5 means $1.50 CPM.
PostConfigPublish
ChatGPTconfig/publish - Publish a draft configuration as the live config. Requires editor access (the configuration:write permission for API keys, or an admin or member workspace role for signed-in callers); viewers are denied. The draft is validated with the same publish-time checks used by the CloudX app before it becomes live.
PostManagedPricingMode
ChatGPTSet dynamic pricing mode - Switch an ad unit between manual and prepared dynamic pricing. A serving dynamic pricing A/B program requires its current report and allocation version to stop safely. Requires configuration:write and editor access.
PostManagedPricingPreview
ChatGPTPreview dynamic pricing setup - Preview the system-selected catalog and traffic split. Set require_baseline to start a dynamic pricing versus line-item A/B comparison. Requires configuration:write and editor access.
PostManagedPricingStart
ChatGPTStart dynamic pricing - Recheck a preview token and live config revision, then publish dynamic pricing and its optional A/B comparison atomically. Requires configuration:write and editor access.
RunScheduledAgent
ChatGPTtasks/run - Run an enabled Task now. Requires editor access and the configuration:write permission.
SearchDocs
ChatGPTSearch CloudX documentation for pages matching the query.
UpdateScheduledAgent
ChatGPTtasks/update - Update a Task. Requires editor access and the configuration:write permission.
managed_pricing_availability
ChatGPTList account ad units with dynamic pricing and A/B eligibility, reasons, mode, and unavailable evidence. Preview is the final setup check.
managed_pricing_mode
ChatGPTSwitch a prepared ad unit between manual and dynamic pricing. Stopping a serving dynamic pricing A/B comparison requires its current report ID and allocation version. This publishes the change immediately.
managed_pricing_preview
ChatGPTPreview dynamic pricing setup. Set require_baseline for an A/B comparison with line items. The response includes the exact token and live config revision needed to start.
managed_pricing_start
ChatGPTStart the approved dynamic pricing setup or A/B comparison using the unchanged preview choices, token, and live config revision. This publishes the change immediately.
Capabilities
Writes
