executing-plans

name	executing-plans
description	Use when you have a written implementation plan to execute in a separate session with review checkpoints

Executing Plans

Overview

Load plan, review critically, execute all tasks, report when complete.

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

Note: Tell your human partner that Superpowers works much better with access to subagents. The quality of its work will be significantly higher if run on a platform with subagent support (such as Claude Code or Codex). If subagents are available, use superpowers:subagent-driven-development instead of this skill.

Fallbacks for local project copies: If referenced workflow skills such as using-git-worktrees or finishing-a-development-branch are not available in the current environment, continue with the closest local equivalent instead of blocking. In this repository, that means: verify the current branch/worktree state yourself before editing, then finish by running the planned verification commands plus the local requesting-code-review and empirical-prompt-tuning completion flow. If the repository has no HEAD yet (for example, an initial uncommitted repo), do not block on commit-range review. Define the review scope as the current uncommitted target files, run a file-level static review against the plan/requirements, and report that the no-HEAD fallback path was used.

If branch or worktree creation is blocked by the sandbox, do not stall the task solely for that reason. Record that branch isolation could not be established in the current environment, avoid destructive git operations, and continue with focused file edits plus explicit verification.

The Process

Step 1: Load and Review Plan

1. Read plan file
2. Re-read the current target files and reconcile them against the plan before editing, especially when the plan came from a previous agent or the worktree is already dirty
3. When the plan extends URL extract/fetch logic across services, verify whether any target service uses query-string identifiers (for example ?content=) and preserve them through normalization and tests rather than rebuilding URLs from pathname alone
4. When the plan adds alternate input formats for an existing identifier (for example prefixed forms or shorthand bare ids), verify that extraction, normalization, and handler tests all converge on one canonical internal id so cache keys and error routing remain stable
5. Review critically - identify any questions or concerns about the plan
6. If concerns: Raise them with your human partner before starting
7. If no concerns: Create TodoWrite and proceed

Step 2: Execute 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: Complete Development

After all tasks complete and verified:

• Announce: "I'm using the finishing-a-development-branch skill to complete this work."
• REQUIRED SUB-SKILL: Use superpowers:finishing-a-development-branch
• Follow that skill to verify tests, present options, execute choice
• If that skill is unavailable, do not stop the task solely for that reason. Run the project's explicit verification commands, perform code review, report results, and note that the fallback path was used.
• If the repository has no HEAD, explicitly say that range-based review is unavailable, review the current uncommitted target files instead, and treat that as the completion-review checkpoint.

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
• Never start implementation on main/master branch without explicit user consent

Integration

Required workflow skills:

• superpowers:using-git-worktrees - REQUIRED: Set up isolated workspace before starting
• superpowers:writing-plans - Creates the plan this skill executes
• superpowers:finishing-a-development-branch - Complete development after all tasks