Task decomposition: big job into executable steps

Your first line in any agent design is what the agent must not do. Most failures come from unclear scope, sloppy tool contracts and missing failure handling, not from a weak model.

Task: An agent must research and write a report, and it wanders off halfway.

[Before you start, do three things]
1. Restate the goal in one sentence: who reads the output, and what decision it serves.
2. If I have not given you the following, ask me. Do not assume:
   - deliverable format and length
   - available sources
   - time and token budget
   - which step needs human sign-off
3. Say how you plan to do it (at most 5 steps) and wait for my go-ahead.

[Go through each of these]
- split into five to eight steps with clear inputs and outputs
- each step states how to know it is done
- mark parallel versus sequential steps
- set checkpoints; stop and report when over budget

[Process]
1. Write the boundary: what it may do, must never do, and when to hand back to a human
2. Define each tool's inputs, outputs, failure semantics, idempotency
3. Write the flow as a state machine with exit conditions
4. Design failure paths: retry, degrade, abort, escalate
5. Give one testable success criterion

[Output]
Fixed shape: boundary, tool contracts, state machine, failure paths, criteria.

[Do not]
- No general-purpose agent that does everything
- Never let fuzzy prose stand in for a tool contract
- The agent must not decide whether approval is needed
- No 'do your best' instead of an abort condition