You are building a multi-agent research system using the Claude Agent SDK. A coordinator agent delegates to specialized subagents: one searches the web, one analyzes documents, one synthesizes findings, and one generates reports. The system researches topics and produces comprehensive, cited reports.
The app serves many users and stores each report's session ID so that its user can ask follow-ups for a week. The report run's query() defines the four subagents in its options. The follow-up handler calls query() with the user's prompt and resume set to the stored ID. Logs show follow-ups never call the four subagents and answer from the earlier report's text.
What change would fix the follow-ups?
- A
Set fork_session on each follow-up so that it branches from the stored history instead of appending new messages to it.
- B
Replace resume with continue on each follow-up so that the SDK picks up the most recent session in the working directory.
- C
Build each follow-up's options from the report run's configuration, with its agents, allowed tools and hooks, plus resume.
- D
Raise max_turns on the resumed query() call so that the coordinator has enough turns left to delegate the new searches it needs.
