babysit-pr
Use after a pull request is open to watch its CI and review-bot comments until it is ready. Fixes real findings, replies with a…
Have an interactive discussion about a topic, approach, feature, or bug without making code changes. Researches the codebase as needed and talks through options. Use to think through an approach before planning or coding.
$ npx -y skills add greenfield-inc/Pane --skill discussion --agent claude-codeHow it fires
How this skill gets triggered: by you, by Claude, or both.
/discussionContext preview
The summary Claude sees to decide when to auto-load this skill.
Have an interactive discussion about a topic, approach, feature, or bug without making code changes. Researches the codebase as needed and talks through options. Use to think through an approach before planning or coding.
name: discussion description: Have an interactive discussion about a topic, approach, feature, or bug without making code changes. Researches the codebase as needed and talks through options. Use to think through an approach before planning or coding. argument-hint: "[topic or question to discuss]"
Talk the topic through with the user, back and forth, until the ideas, tradeoffs, and goal are clear enough to plan or build.
Read code and research freely. Your only output is conversation. Never edit, create, or delete project files, and never propose diffs or patches to apply.
Use the `explorer` subagent or `codebase-explorer` for code, and `research-web` for external libraries or approaches, including mid-conversation when new questions come up.
it.
tradeoffs.
and defer to the user's judgment.
Classify every fork question before you raise it. If running something can answer it (behavior, timing, output, performance, layout), build the cheapest throwaway probe outside the project tree, run it, and present the result with a recommendation. Save questions for product or preference calls that no experiment can settle. A probe usually answers faster than a person, and it gives them a result to react to.
Recap in the conversation:
Mark which points are the user's intent and which are your assumptions. Then recommend the next step, usually `create-ticket` when the work is ready to delegate, `options` when more than one approach is live, `brief` for a write-up, `investigate` for a bug, or implementation.
Suggested next steps: - `/create-ticket [description]` - Capture the work for delegation - `/options [question]` - Lay out the approaches and trade-offs - `/discussion [follow-up]` - Continue exploring a specific aspect - `/research-web [topic]` - Look up external documentation
Topic to discuss: $ARGUMENTS
Use after a pull request is open to watch its CI and review-bot comments until it is ready. Fixes real findings, replies with a…
Use when a problem, idea, or change of intent is worth capturing as a long-form page, or the person asks to write it up. Never…
Use when the person reports broken behaviour or starts a session to file a bug. Reproduces and reports. Never fixes.
Shared vocabulary for designing deep modules. Use when the user wants to design or improve a module's interface, find deepening…
Explore the codebase to locate files, trace behavior, and document existing patterns with precise file references. Use when the…
Give a fresh agent something to walk through with zero context - a PR body, a landing page, a document, a process, instructions,…