Load plan, review critically, execute all tasks, report when complete. For inline execution without subagents. If subagents are available, use bp:subagent-driven-development instead — quality is significantly higher.
Use when you have a written implementation plan and the user chose inline execution.
Use when subagents are unavailable but a plan needs executing.
Announce at start: "I'm using the executing-plans skill to implement this plan."
**Load and review** — Read plan. Review critically for questions or concerns. If concerns: raise with human partner before starting. If clear: create TodoWrite and proceed.
**Execute tasks** — For each task: mark in_progress, follow steps exactly (plan has bite-sized steps), run verifications as specified, mark completed.
**Complete development** — After all tasks verified, invoke bp:finishing-a-development-branch.
Follow plan steps exactly — don't skip verifications.
Stop when blocked — don't guess. Ask for clarification.
Never start implementation on main/master without explicit user consent.
STOP executing immediately when: hit a blocker (missing dep, test fails, instruction unclear), plan has critical gaps, or verification fails repeatedly. Ask for help rather than guessing.
Integration
- bp:using-git-worktrees — set up isolated workspace before starting
- bp:writing-plans — creates the plan this skill executes
- bp:finishing-a-development-branch — complete development after all tasks