[FEATURE] Per-agent effort level control

Status Closed — duplicate
Maintainer reply None cached
Activity 3 comments · opened Apr 13, 2026 · closed Apr 16, 2026

When spawning subagents via the Agent tool, it would be useful to set an effort parameter (e.g., low/medium/high) per agent independently of the main conversation's effort level.

Use case: Run quick lookup agents at low effort (fast, cheap) while keeping the main conversation at high effort for complex work. This would improve parallel workflows where not every agent needs full reasoning depth.

View original on GitHub ↗

This issue has 3 comments on GitHub. Read the full discussion on GitHub ↗