feat: /plan-pm-review — RICE prioritization, JTBD segmentation, acceptance criteria (fixes #1412)

[NikhileshNanduri]

Summary

Closes #1412.

Adds /plan-pm-review, the missing PM voice in gstack's plan review pipeline, and wires it into /autoplan as Phase 1.5.

Three modes:

• PRIORITIZE — RICE scoring (Reach × Impact × Confidence / Effort) on every plan item. Produces a ranked cut list, a risk-adjusted confidence table for low-confidence items, and a dependency-ordered P1 roadmap. Effort is expressed in CC+gstack time (not human-team months), so scores reflect AI-compression reality.
• SHARPEN — Turns vague plan items into acceptance criteria with testable done conditions. A done condition is valid only if a QA engineer can verify it without asking the developer a question.
• SEGMENT — JTBD analysis: functional job, emotional job, social job, firing trigger, hiring trigger, success metric. Produces a segment-to-feature coverage matrix and explicit misalignment flags.

All three modes produce a "NOT in scope" section (formal deferrals) and a "What already exists" section (prevents feature duplication).

/autoplan Phase 1.5:

• Inserted between Phase 1 (CEO) and Phase 2 (Design)
• Skip conditions: plan ≤ 3 deliverables AND segment already named AND no TODOS conflicts
• Run conditions: >5 work items OR multiple user segments OR borderline scope OR TODOS conflicts
• Auto-selects mode (PRIORITIZE by default) — no AskUserQuestion interruptions in autoplan flow

Changes

• plan-pm-review/SKILL.md.tmpl — new skill template (378 lines)
• plan-pm-review/SKILL.md — generated from template
• scripts/resolvers/tasks-section.ts — added 'pm-review' to VALID_PHASES
• autoplan/SKILL.md.tmpl — Phase 1.5, Pre-Gate Verification, Review Scores, Completion logs
• autoplan/SKILL.md — regenerated
• AGENTS.md, docs/skills.md — documentation
• VERSION: 1.43.3.0 → 1.44.0.0
• CHANGELOG.md: v1.44.0.0 entry

Test plan

• bun test passes — 718 tests, 0 failures (skill validation + gen-skill-docs)
• bun run gen:skill-docs generates plan-pm-review/SKILL.md cleanly at ~17K tokens
• {{TASKS_SECTION_EMIT:pm-review}} resolves without throwing
• Doc inventory cross-checks pass: skill appears in AGENTS.md and docs/skills.md