Skip to content

Enhanced ChatGPT Clone: Features Agents, MCP, Skills, DeepSeek, Anthropic, AWS, OpenAI, Responses API, Azure, Groq, o1, GPT-5, Mistral, OpenRouter, Vertex AI, Gemini, Artifacts, AI model switching, message search, Code Interpreter, langchain, DALL-E-3, OpenAPI Actions, Functions, Secure Multi-User Auth, Presets, open-source for self-hosting. Active

2.8 KB
Raw Blame

Scheduled chats: deferred scope

Scheduled chats remain experimental and default-off.

Must-have product follow-ups

  • Add one-time and monthly schedules; current cadences are hourly, daily, weekdays, and weekly.
  • Add non-agent chat targets and existing-conversation continuation; current schedules require agent_id and always target new.

Client/API parity

  • Return minIntervalMinutes in schedule limits and surface it in the dialog.
  • Add active-run and run-history discovery. The list response currently projects only lastRun, which is written on pause, skip, or terminal outcomes, so the client's started chip is unreachable.
  • Add multi-day weekly, timezone, and attachment controls. The API supports these, but the dialog does not. NOTE: the dialog cannot scroll from md up (md:overflow-visible exists because the Ariakit popovers cannot portal out of the focus-trapping Radix dialog), so its content must fit the viewport — roughly 30px of slack remain at 1280x720. Each of these controls adds a row and will push the footer's Save button out of reach. Adding them needs the height problem solved first: give ControlCombobox the portalElement prop Dropdown already has, portal the popovers into the dialog content element, and let the form scroll again.

Known gaps: project scope

  • A project deleted inside the window between the resume controller's isScheduleLive(..., { policy: true }) check and claimScheduleResume is not caught: deletion bumps no revision, so the lease fence cannot see it and the continuation starts against a conversation that was just unscoped. Sub-second, and self-correcting at the next fire (which auto-disables the schedule). Closing it needs a distinct policy conflict returned from the claim that the controller routes through abort-and-settle, rather than the bare 409 an inactive conflict produces today.
  • The stored project converges on an operator pin only when the schedule actually FIRES. A schedule that has not fired since the pin moved still carries its old id, which the wire projection already papers over but a direct read of the row does not. Resume validation is unaffected — it reads the occurrence's own record.

Runtime scope

  • The standard server supports multi-replica scheduling only with a confirmed Redis-backed GenerationJobManager; in-memory mode requires SCHEDULES_SINGLE_PROCESS. The legacy experimental.js clustered entrypoint does not arm the engine and rejects schedule writes.
  • Deferred account deletion blocks new remote API admission but cannot yet drain already-admitted OpenAI-compatible or Responses requests; see #14594.

Test debt

  • Add Playwright coverage for creating a schedule through the UI. The execution E2E currently seeds schedules through the API.