implementlisted

# Executing a plan **Speak the repository's language.** The first script this skill runs — `<cue>/gate`, or `<cue>/init-check` — prints a `language` line. **Everything you say to the user goes in that language, starting with your first sentence**: the questions you ask, not only the record you write at the end. This skill body is English because it is the plugin's source, not because it is your output. Template headings, commit markers, `outcome.md` status tokens, paths and code stay exactly as written. Turn an approved plan into commits, and leave a record that says honestly what happened — including where it diverged and what was left over. **Announce at start:** "I'm using the cue-dev:implement skill to execute the plan." **Narration:** between tool calls, narrate at most one short line — the records and the tool results carry the rest. **Continuous execution:** Do not pause to check in between tasks. The only reasons to stop are a blocker you cannot resolve, ambiguity that genuinely prevents progress, or all tasks complete. "Should I continue?" wastes your human partner's time — they asked you to execute the plan. ## This file does not contain the method There are two ways to execute a plan and this item has already been assigned one. **The procedure lives in its own file, and you read exactly one of them:** | the item's level | the procedure | |---|---| | `independent-review` | [independent-review.md](independent-review.md) | | `self-review` | [self-review.md](s