Purpose
Produce a ready-to-use set of open, non-leading coaching questions calibrated to a specific people challenge and relationship — structured through the GROW model and layered with scenario-specific probes — so an HR partner or people manager walks into a difficult conversation with the right questions instead of defaulting to advice-giving or avoidance.
When to Use
Invoke this skill when the user asks to:
- Prepare for a coaching conversation about an underperforming employee
- Get questions to help someone work through a conflict with a colleague or manager
- Coach a team member who is resisting a change or restructure
- Support a career development or growth conversation
- Coach a manager on a capability or leadership gap
- Build a question bank for a recurring coaching situation
Inputs Required
Before starting, collect:
- The people challenge: What specifically is happening (underperformance, conflict, change resistance, career stall, manager capability, etc.)
- Who is being coached: Role, seniority, and relationship to the coach (direct report, peer, someone else's report)
- What outcome the coach wants: What "better" looks like after the conversation
- What's been tried: Any prior conversations, feedback, or interventions
- Emotional temperature: How sensitive or charged the situation is (routine / delicate / high-stakes)
- Coaching model preference: GROW / OSKAR / CLEAR — default to GROW if unspecified
- Time available: 15 min check-in / 30–45 min full session
If the coaching model or time is missing, default to GROW and a 30–45 minute session, and note the assumption.
Instructions
Step 1: Frame the Coaching Stance
Before generating questions, set the operating principle: the coach's job is to help the person think, not to deliver a verdict. Confirm the challenge is genuinely coachable (a skills/mindset/motivation gap) versus a matter requiring a directive conversation (misconduct, safety, non-negotiable policy). If it's the latter, flag it — coaching questions are the wrong tool for undebatable expectations.
Step 2: Generate 45 Questions Across the GROW Model + Scenario Probes
Generate every question fully written for this specific challenge and relationship — not generic templates. Label each with the intent it serves.
Goal — What does the person want? (8 questions) Surface the person's own definition of success and what matters to them.
- Draw out their picture of a good outcome, not the coach's
- Include at least one question that separates the presenting problem from the deeper goal
- Include one scaling question ("On a scale of 1–10...")
Reality — What's actually happening now? (10 questions) Build a shared, honest view of the current situation without judgment.
- Include questions that surface facts, patterns, and the person's own role in the situation
- Include one question that gently checks their assumptions
- Include one question inviting perspective-taking (how others might see it)
Options — What could the person do? (8 questions) Expand the range of possibilities before narrowing.
- Push for quantity of options before evaluating any
- Include one question that removes a constraint ("If [obstacle] weren't a factor...")
- Include one question drawing on their past success in a similar spot
Will / Way Forward — What will the person commit to? (7 questions) Convert insight into a concrete, owned next step.
- Include a specific-first-action question
- Include a commitment-strength check and an accountability question
- Include one that surfaces what support they need
Scenario-Specific Probes (12 questions) Tailored to the exact challenge provided. Match the set to the situation:
- Underperformance: questions distinguishing capability from motivation from clarity; questions surfacing blockers the person hasn't named
- Conflict: questions inviting the other party's perspective, separating position from need, exploring what a repaired relationship enables
- Change resistance: questions naming the loss underneath the resistance, surfacing what they'd need to feel confident, connecting the change to something they value
- Career development: questions exploring energy vs. skill, expanding beyond the obvious next title, surfacing what's holding them back
- Manager capability: questions shifting focus from the task to the team's experience, surfacing their model of what a good manager does
Step 3: Build Follow-Up Sequences
For 3 pivotal questions, show a 2–3 question follow-up chain — how to go deeper when the first answer is surface-level, defensive, or vague. Model the "and what else?" and "say more about that" moves.
Step 4: List Questions to Avoid
Provide 5 leading, judgmental, or closed questions specific to this scenario that undermine coaching — each paired with a reworded open version. Include the classic trap: "Why did you...?" (blame-triggering) reworked to "What was happening when...?"
Step 5: Suggest a Conversation Flow
Provide a suggested order and time allocation for the session, mapping which question sets to open with, when to move between GROW phases, and how to close on commitment.
Output Format
- Structure: GROW phases as labeled sections with numbered questions; separate sections for scenario probes, follow-up sequences, questions to avoid, and conversation flow
- Length: 1,000–1,400 words
- Format: Scannable — numbered question lists with brief intent labels, prose only for the flow and stance sections
- Sections: Coaching stance note → Goal → Reality → Options → Will → Scenario-specific probes → Follow-up sequences → Questions to avoid → Suggested conversation flow
- Tone: Calm, curious, non-directive — the language a skilled coach actually uses, not corporate HR jargon
Quality Checklist
Before delivering output, verify:
- [ ] All 45 questions are present (count them: 8 + 10 + 8 + 7 + 12)
- [ ] Every question is open (starts with what/how, not why/did/should) — except deliberate exceptions that are labeled
- [ ] Questions are written for the specific person and challenge — no bracket placeholders or generic filler
- [ ] Scenario-specific probes actually match the challenge provided
- [ ] Follow-up sequences show a genuine deepening move, not a rephrase
- [ ] Each "question to avoid" is paired with a working reworded version
Notes
- The most common failure in coaching conversations is smuggling advice inside a question ("Have you thought about just...?") — the generated set should model genuinely open inquiry
- "And what else?" is the single most useful follow-up in coaching — it consistently surfaces the real answer after the rehearsed one
- If the situation involves misconduct, legal risk, or a non-negotiable expectation, name it clearly: that requires a direct conversation, and coaching questions can follow only once the expectation is unambiguous
- Scaling questions (1–10) are useful across every scenario because they turn abstract feelings into a workable starting point
How to use this skill
This file is a Claude Code skill in Anthropic's standard format — markdown with YAML frontmatter at the top.
Installing in Claude Code:
~/.claude/skills/<folder-name>/SKILL.md— auto-loads in every session anywhere on your machine (recommended for skills you'll use across multiple projects)<your-project>/.claude/skills/<folder-name>/SKILL.md— auto-loads only when working in that project (recommended for project-specific skills)
Restart Claude Code or open a new session to register.
Using in Claude.ai, ChatGPT, Gemini, or other AI tools:
- Drag this file into a new chat, or copy and paste its contents
- Send your task or request as your next message — the AI will apply the skill where it fits
For Claude Code skill details: https://docs.anthropic.com/en/docs/claude-code/skills