executing-plans

Executing Plans

Overview

Load plan, review critically, execute all tasks directly in the current session, and report completion.

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

Branch policy (required): Execute only on main or master. Do not create feature branches, do not use git worktrees, and do not dispatch subagents.

Implementation policy (required):

1. Read the plan **Architecture:** line. It must indicate Lightweight or Heavyweight and which primary skill applies (lightweight-software-principles or hex-infra).
2. If the plan is missing a clear approach (no Approach=Lightweight|Heavyweight or equivalent), stop and run superpowers:selecting-implementation-architecture with the user, then update the plan header before continuing.
3. Always use superpowers:test-driven-development for RED-GREEN-REFACTOR.
4. Conditionally use superpowers:lightweight-software-principles (lightweight) or superpowers:hex-infra (heavyweight) for structure and boundaries — as recorded in the 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. Confirm current branch is main or master (git branch --show-current)
4. Parse **Architecture:** for Lightweight vs Heavyweight; if missing or ambiguous, run superpowers:selecting-implementation-architecture and reconcile the plan
5. If concerns: Raise them with your human partner before starting
6. If no concerns: Create TodoWrite and proceed

Step 2: Execute Tasks

For each task:

1. Mark as in_progress
2. Apply the plan's architecture skill:
   • Heavyweight: Classify command vs query; follow superpowers:hex-infra (handlers, ports, adapters, domain craftsmanship)
   • Lightweight: Follow superpowers:lightweight-software-principles (modular, domain-centered, thin edges); classify command vs query only when it clarifies the change
3. Follow each step exactly (plan has bite-sized steps)
4. Run verifications as specified
5. Mark as completed

Step 3: Complete Development

After all tasks complete and verified:

• Run full project verification one more time
• Summarize exactly what changed
• Ask whether to commit the changes to main/master now or leave them uncommitted

When to Stop and Ask for Help

STOP executing immediately when:

• Hit a blocker (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.

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
• Stop when blocked, don't guess
• Always execute on main/master
• Never create worktrees or feature branches as part of this flow
• Never use subagents for this flow
• Follow the architecture recorded in the plan (lightweight-software-principles or hex-infra)

Integration

Required workflow skills:

• superpowers:writing-plans - Creates the plan this skill executes (includes architecture selection)
• superpowers:selecting-implementation-architecture - Use when the plan lacks a clear Lightweight vs Heavyweight decision
• superpowers:lightweight-software-principles - Use when plan Architecture specifies Lightweight
• superpowers:hex-infra - Use when plan Architecture specifies Heavyweight
• superpowers:test-driven-development - Always follow RED-GREEN-REFACTOR while implementing
• superpowers:requesting-code-review - Review completed work before final handoff