plan-execute

plan-execute separates thinking from doing: first create a structured plan, then execute each step, then verify the result. If verification fails, the loop repeats.

When to use it

• Complex tasks that benefit from upfront planning
• Multi-step workflows with verifiable outcomes
• Tasks where the execution order matters
• Code generation and testing loops

File structure

agents/my-plan-execute/
├── AGENT.md    ← pattern: plan-execute, config
├── execute.md  ← execute one step from the plan
└── verify.md   ← verify the current result

AGENT.md

---
name: project-builder
description: Plans and executes a multi-step project.
version: "1.0.0"
pattern: plan-execute
max_iter: 5
call:
  model:
    role: thinker
---

You are a project planner. Given a task, create a detailed execution plan.

Output a JSON array of steps:
[
  {"step": 1, "action": "...", "expected_output": "..."},
  {"step": 2, "action": "...", "expected_output": "..."}
]

Call finish with key="done" and the JSON plan as value.

execute.md

---
call:
  model:
    role: thinker
  uses:
    - search-web
---

You receive a plan and the current execution state.

Execute the next pending step. Record what you did and what you produced.

Update the plan state to mark the step as complete.

Call finish with key="done" and the updated execution state as value.

verify.md

You receive the completed execution state.

Check:
1. All steps were executed
2. Each step's output matches its expected output
3. The overall goal is achieved

If everything is complete and correct, call finish with key="DONE".
If steps need to be retried or the plan needs adjustment, call finish with key="RETRY" and describe what needs to change.

Scaffold

tama add plan-execute my-agent

How it works

AGENT.md body → create JSON plan
      ↓
execute.md → execute next step, update state
      ↓
verify.md → check results
  → key="DONE": return final result
  → key="RETRY": loop back to execute
      ↓
(repeat up to max_iter times)

DONE and RETRY

The verify step controls the loop:

• finish(key="DONE", value=<result>) — all steps complete, return result
• finish(key="RETRY", value=<feedback>) — something failed, loop back with feedback