[scarletkc]

Users with an unclear idea need help deciding what they want before they can provide an implementation brief. Add ask-to-plan, a skill that uses the harness's native questions and choices to narrow goals, requirements, constraints, solution, and technology choices into an actionable plan.

The interview adapts to answers, supports uncertainty and corrections, distinguishes pending input from a user decision, and stops when the plan is actionable. It follows the host's tool availability and input restrictions. Add the skill to the generated README catalog.

Validation:

• uv run --python 3.12 python -m unittest discover -s tests -v — 77 tests passed.
• uv run --python 3.12 python scripts/update_readme_skills.py --check — passed.
• Skill Creator quick_validate.py on skills/ask-to-plan — passed.
• gh skill publish --dry-run — passed.
• git diff --cached --check — passed before commit.
• Manual native-input smoke test covered choice selection, free-text context, adaptive follow-up questions, and revising a previous answer. Final plan delivery was not part of that smoke test.