Executing Plans
Run a saved implementation plan task-by-task with review checkpoints, verification steps, and a defined handoff when the branch is ready to finish.
Overview
Executing-plans is an agent skill most often used in Build (also Ship) that loads a written implementation plan, executes each task with verifications, and finishes the branch through a structured completion workflow.
Install
npx skills add https://github.com/obra/superpowers --skill executing-plansWhat is this skill?
- Loads a written plan, critiques it up front, and stops for human input when the plan is unclear or risky
- Executes bite-sized tasks in order with in-progress/completed tracking and per-task verifications from the plan
- Ends by routing through finishing-a-development-branch for tests, options, and clean branch completion
Adoption & trust: 109k installs on skills.sh; 221k GitHub stars; 3/3 security scanners passed (skills.sh audits).
What problem does it solve?
You have a detailed implementation plan but need a disciplined agent session that executes tasks in order, verifies each step, and stops cleanly when something blocks progress.
Who is it for?
Solo builders running a second session dedicated to implementing an existing plan file with explicit checkpoints and stop-on-blocker rules.
Skip if: Teams without a written plan yet, one-off chat coding without task lists, or environments where you should use subagent-driven-development instead of sequential execution.
When should I use this skill?
Use when you have a written implementation plan to execute in a separate session with review checkpoints.
What do I get? / Deliverables
You get a fully worked implementation with completed plan tasks, passing verifications where specified, and a branch finished via the Superpowers development-branch closing ritual.
- All plan tasks marked complete with verifications run as specified
- Blocked or ambiguous items escalated to the human before further execution
- Branch completion handled through finishing-a-development-branch
Recommended Skills
Journey fit
Spans multiple journey phases - primary shelf plus alternate fits below.
Canonical shelf is Build because the skill’s job is turning an existing implementation plan into shipped code changes, not ideation or distribution. PM subphase fits plan-driven execution: load the plan, track tasks, and drive the build session as structured project work rather than a single integration or UI slice.
Where it fits
Execute a thin prototype plan in one session so spikes stay scoped to listed files and verifications.
Implement API and data-layer tasks from a plan without skipping ordered migrations and integration checks.
Work through UI tasks with the plan’s component and route steps plus explicit manual or test verifications.
Rely on per-task verification commands and the finishing sub-skill so failing tests halt execution before merge.
How it compares
Use instead of asking the agent to improvise implementation from memory when you already exported a step-by-step plan from writing-plans.
Common Questions / FAQ
Who is executing-plans for?
Executing-plans is for solo and indie builders using Claude Code, Codex, Cursor, or similar agents who keep implementation plans in the repo and want repeatable, checkpointed execution sessions.
When should I use executing-plans?
Use it in Build when you have a finished plan to implement task-by-task; use it in Ship when the plan’s verifications and the required finishing-a-development-branch step are how you close testing and review before merge or deploy.
Is executing-plans safe to install?
It orchestrates normal development work including shell, git, and filesystem actions implied by your plan—review the Security Audits panel on this Prism page and inspect the skill package in your repo before granting agent permissions.
SKILL.md
READMESKILL.md - Executing Plans
# 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. ## 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 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 ## 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** - Ensures isolated workspace (creates one or verifies existing) - **superpowers:writing-plans** - Creates the plan this skill executes - **superpowers:finishing-a-development-branch** - Complete development after all tasks