Executing Plans

Executing Plans

Overview

Load plan, review critically, execute tasks in batches, report for review between batches.

Core principle: Batch execution with checkpoints for architect review.

Announce at start: "I'm using the executing-plans skill to implement this plan."

The Process

Step 1: Load and Review Plan

1. Read plan file
2. Review critically - identify any questions or concerns about the plan
3. If concerns: Raise them with your human partner before starting
4. If no concerns: Create TodoWrite and proceed

Step 2: Execute Batch

Default: First 3 tasks

For each task:

1. Mark as in_progress
2. Follow each step exactly (plan has bite-sized steps)
3. Run verifications as specified
4. Mark as completed

Step 3: Report

When batch complete:

• Show what was implemented
• Show verification output
• Use AskUserQuestion for feedback:

Use AskUserQuestion tool with:
- question: "Batch complete. How should I proceed?"
- header: "Feedback"
- options:
  - label: "Continue"
    description: "Looks good, proceed to next batch"
  - label: "Revise"
    description: "I have feedback on this batch first"
  - label: "Pause"
    description: "Stop here, I'll review more carefully"
- multiSelect: false

Step 4: Continue

Based on selection:

• Continue: Execute next batch
• Revise: Wait for feedback, apply changes, then continue
• Pause: Stop execution, preserve progress

Step 5: Complete Development

After all tasks complete and verified:

• Announce: "I'm using the finishing-a-development-branch skill to complete this work."
• Use finishing-a-development-branch skill
• Follow that skill to verify tests, present options, execute choice

Step 6: Next Action Loop (CRITICAL - Issue #118)

After completion, ALWAYS present next actions:

Use AskUserQuestion tool with:
- question: "What would you like to do next?"
- header: "Next Action"
- options:
  - label: "Work on another issue"
    description: "Continue with another GitHub issue"
  - label: "Run tests/checks"
    description: "Run test suite or quality checks"
  - label: "Session capture and exit"
    description: "Save state for later continuation"
- multiSelect: false

NEVER end execution without presenting next step options.

If user selects "Work on another issue", fetch prioritized open issues:

gh issue list --state open --milestone v1.0.0 --json number,title,labels --limit 5

Then present specific issue options via AskUserQuestion.

When to Stop and Ask for Help

STOP executing immediately when:

• Hit a blocker mid-batch (missing dependency, test fails, instruction unclear)
• Plan has critical gaps preventing starting
• You don't understand an instruction
• Verification fails repeatedly

Ask for clarification rather than guessing.

PDF Input Support

Plans can be provided as PDF files:

User: Execute this plan: /path/to/implementation-plan.pdf

Process PDF plans:

1. Use Read tool to analyze the PDF content
2. Extract tasks, steps, and verification criteria
3. Convert to internal task list format
4. Proceed with standard execution process

When reading plan PDFs:

• Look for: numbered tasks, code blocks, file paths
• Extract: exact commands, expected outputs
• Note: dependencies between tasks
• Identify: verification steps for each phase

PRD PDFs can also be processed to understand requirements context before execution.

When to Revisit Earlier Steps

Return to Review (Step 1) when:

• Partner updates the plan based on your feedback
• Fundamental approach needs rethinking

Don't force through blockers - stop and ask.

Remember

• Review plan critically first
• Follow plan steps exactly
• Don't skip verifications
• Reference skills when plan says to
• Between batches: just report and wait
• Stop when blocked, don't guess