← ClaudeAtlas

research.implementlisted

Work the tasks.md queue across Setup, Build, Eval, explicit user-led manuscript work, and Polish; keep claims and task status current.
jiancui-research/research-kit · ★ 3 · AI & Automation · score 75
Install: claude install-skill jiancui-research/research-kit
> **research-kit stage - `/research.implement`.** > > The instructions for this stage are not duplicated here. Read > `commands/research.implement.md` from this plugin's own directory - the folder two levels above > this file - and follow it end to end. If you cannot find it, say so and stop rather > than reconstructing the stage from memory. > > Two adaptations from its original slash-command form: > > - Where it references `$ARGUMENTS`, that means the user's latest message to you - > their free-text input for this stage (task id/section (e.g. T012, eval, or "paper intro"); Manuscript mode also accepts outline, critique, or explicit "draft <section>). When they gave none, follow the > step's "if empty" guidance or ask for it; never invent one. > - Where a step ends with `Next: /research.<x>`, run the `research.<x>` skill next. > > Everything else is unchanged: read and write only under `./.research/`, follow the > command contract, and stay paper-type aware.