Create a Discovery Questionnaire for an Expert

What it does

This workflow interviews the sender about the recipient and the decision they need to make, rather than demanding answers the sender does not know. The output explains its purpose, context, expected effort and use of responses. Questions are ordered by importance, grouped by theme when helpful and limited to one idea each. Answer stubs invite partial responses and explicit uncertainty. The source writes a local to-questionnaire-topic.md file; creating it is distinct from sending it to the recipient.

When to use it

• Ask a service owner for constraints needed to plan an integration.
• Collect domain facts from a colleague unavailable for a live interview.
• Prepare a meeting document that links questions to a concrete decision.

Step-by-step workflow

1. Clarify the recipient’s role, expertise and relationship to the sender.
2. List the decisions or facts the sender needs back.
3. Write enough context for someone outside the original conversation.
4. Draft prioritized single-idea questions with answer space and uncertainty guidance.
5. Save the Markdown file and check that every required decision is covered.

Example request

Use to-questionnaire to prepare questions for our warehouse operations lead. I need enough information to decide whether returns can be processed overnight. Ask me about the recipient and intended decision, not the warehouse facts I lack. Include a short context section, realistic response effort and space for unknown answers. Save locally; do not email or message anyone.

Worked example

An overnight-returns questionnaire should ask about cutoff times, staffing and exceptions separately. Combining all three in one question makes partial answers difficult to interpret. The recipient needs to know whether answers will guide a pilot or a permanent staffing decision. If no deadline was agreed, leave it explicit instead of inventing urgency. The closing catch-all gives the expert a way to reveal an important constraint the sender could not anticipate.

What you get

A local discovery questionnaire whose answers map to the sender’s stated decisions.

How to evaluate the result

• Map every required decision to at least one question.
• Check the recipient can understand the purpose without prior chat context.
• Verify unknown and partial answers are explicitly welcome.

What you need

A defined recipient and a clear account of what the sender needs to learn. Sensitive business details should be included only when appropriate for the eventual recipient.

Setup and activation

The publisher documents plugin setup for Claude Code, Codex and Copilot, and manual Agent Skills setup for Gemini CLI and other clients. Choose one distribution method per client to avoid duplicate skills. The engineering collection expects project-specific tracker, triage-label and document-location settings from setup-matt-pocock-skills. Keep referenced companion skills and files available. This catalogue review does not install or run them; scan the exact source before any later installation. Client tools, permissions and subagent support still need separate validation.

Limits and considerations

The document does not verify the expert’s answers or grant permission to send it. One recipient’s knowledge may leave gaps. Avoid fabricated deadlines, compound questions and collecting unnecessary personal information.

Related workflows

• Use grilling when the decision owner can answer live.
• Use handoff to transfer the returned answers into another work session.

License and reuse

MIT at repository level (Copyright 2026 Matt Pocock). Preserve the copyright and permission notice when redistributing substantial source material; inspect any separately licensed dependencies or assets.

Further reading

• Reviewed SKILL.md (immutable commit)
• Publisher README: distributions and compatibility
• Reviewed MIT license