mnhat-exploring

name	mnhat-exploring
description	Capture locked decisions for ambiguous work before planning. Use when a feature, bug fix, or workflow request still needs scope, boundaries, and a canonical `context.md` before implementation planning.

MNHAT Exploring

Quick start

• Ask one question at a time.
• Lock only user-facing and scope-defining decisions.
• Write workstream artifacts under .mnhat/<YYYY-MM-DD>-<feature-slug>/exploring/.
• Stop when downstream planning can proceed without guesswork.

Workflows

1. Classify scope as Quick, Standard, or Deep.
2. Pick only relevant probes from references/gray-area-probes.md.
3. Do a shallow scout of current repo reality.
4. Lock decisions with stable IDs such as D1, D2, D3.
5. Write context.md from references/context-template.md.
6. Write open-questions.md and resolved-questions.md when they have real content.
7. Hand off to mnhat-planning or goal-griller.

Outputs

• Canonical artifacts:
  • .mnhat/<YYYY-MM-DD>-<feature-slug>/exploring/context.md
  • .mnhat/<YYYY-MM-DD>-<feature-slug>/exploring/open-questions.md
  • .mnhat/<YYYY-MM-DD>-<feature-slug>/exploring/resolved-questions.md
• Canonical template owners:
  • references/context-template.md
  • references/open-questions-template.md
  • references/resolved-questions-template.md

State rules

• context.md is the primary artifact when state.json.phase is exploring.
• Do not mutate code, create execution tasks, or write planning artifacts here.
• If a meaningful decision is made that should survive resume, call mnhat-implementation-record.

Further reading

• references/context-template.md - canonical context.md shape
• references/open-questions-template.md - canonical open-questions.md form
• references/resolved-questions-template.md - canonical resolved-questions.md form
• references/gray-area-probes.md - question bank for ambiguous requests