Pipeline integrity + brownfield adoption system

Pipeline integrity (4 fixes for snapshot-vs-live-reference problem):
- NEW docs/architecture/tr-registry.yaml: persistent stable TR-IDs per GDD
  requirement; /architecture-review bootstraps and appends, never renumbers
- /create-control-manifest: added Manifest Version stamp to header
- /create-epics-stories: stories embed TR-ID reference (not quoted text),
  Manifest Version from manifest, ADR status gate (Proposed → Blocked)
- /story-done: TR-ID registry lookup at review time, manifest staleness check
- /story-readiness: ADR Accepted check, TR-ID validity, manifest version check
- /review-all-gdds + /architecture-review: fixed parenthetical status values
  ("Needs Revision" only, no parentheticals that break exact-match reads)

Workflow infrastructure:
- NEW /help skill: context-aware "what's next" using workflow-catalog.yaml
- NEW .claude/docs/workflow-catalog.yaml: YAML-driven phase/step sequence
- /sprint-plan + /sprint-status: sprint-status.yaml machine-written tracking

Brownfield adoption system (migrate, not replace):
- NEW /adopt skill: format compliance audit (not existence check); classifies
  gaps BLOCKING/HIGH/MEDIUM/LOW; produces docs/adoption-plan-[date].md with
  numbered migration plan; never regenerates existing artifacts
- /design-system retrofit [path]: fills only missing GDD sections, preserves
  all existing content
- /architecture-decision retrofit [path]: adds missing ADR sections (Status,
  ADR Dependencies, Engine Compatibility) without touching existing content
- /start option D: split into D1 (early-stage) and D2 (has GDDs/ADRs → adopt)

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

.claude/agent-memory/lead-programmer/MEMORY.md

@@ -0,0 +1,28 @@
+# Lead Programmer — Agent Memory
+
+## Skill Authoring Conventions
+
+### Frontmatter
+- Fields: `name`, `description`, `argument-hint`, `user-invocable`, `allowed-tools`
+- Read-only analysis skills that run in isolation also carry `context: fork` and `agent:`
+- Interactive skills (write files, ask questions) do NOT use `context: fork`
+- `AskUserQuestion` is a usage pattern described in skill body text — it is NOT listed
+  in `allowed-tools` frontmatter (no existing skill does this)
+
+### File Layout
+- Skills live in `.claude/skills/<name>/SKILL.md` (subdirectory per skill, never flat .md)
+- Section headers use `##` for phases, `###` for sub-sections
+- Phase names follow "Phase N: Verb Noun" pattern (e.g., "Phase 1: Find the Story")
+- Output format templates go in fenced code blocks
+
+### Known Canonical Paths (verify before referencing in new skills)
+- Tech debt register: `docs/tech-debt-register.md` (NOT `production/tech-debt.md`)
+- Sprint files: `production/sprints/`
+- Epic story files: `production/epics/[epic-slug]/story-[NNN]-[slug].md`
+- Control manifest: `docs/architecture/control-manifest.md`
+- Session state: `production/session-state/active.md`
+- Systems index: `design/gdd/systems-index.md`
+- Engine reference: `docs/engine-reference/[engine]/VERSION.md`
+
+### Skills Completed
+- `story-done` — end-of-story completion handshake (Phase 1-8, writes story file)