mirror of
https://github.com/HKUDS/nanobot.git
synced 2026-05-20 08:32:25 +00:00
- Expand LongTaskTool.description to instruct parent agent on goal construction, return value semantics, and how to handle results. - Expand CompleteTool.description to emphasize that the summary IS the final answer returned to the parent agent. - Prefix validated return value with an explicit "final answer" directive to stop parent agent from re-running work. - Redesign step_start.md: Step 1 is now explicitly for exploration, planning, and skeleton-building. complete() is discouraged. - Remove bulky payload debug logging from _emit(); add targeted info/warning/error logs at key state transitions instead. - Add signal_type to HandoffState for cleaner signal detection.
1.0 KiB
1.0 KiB
Long Task — Step {{ step + 1 }}/{{ max_steps }}
You are the FIRST step in a meta-ReAct loop. Your job is NOT to finish the entire goal — it is to explore, plan, and build the skeleton.
Goal
{{ goal }}
Instructions
- Explore: Check the filesystem and any relevant state. Do NOT assume anything.
- Plan the split: Explicitly decide how the goal will be divided across the remaining {{ max_steps - 1 }} steps. Write this plan down.
- Do ONE chunk: Make concrete progress on ONLY the first chunk. Do NOT attempt to finish the entire goal now, even if you have enough tool calls. The meta-loop exists so later steps can review, correct, and refine your work.
- Handoff: Call
handoff()with:- A detailed summary of what you did
- Files changed
- The explicit plan for the remaining steps
- A clear hint for the next step
You have {{ budget }} tool calls. Reserve the last 1-2 for handoff().
Do NOT call complete() in Step 1 unless the goal is literally a single trivial action.