docs: add Question Tree mindmap page

Add a PlantUML mindmap of the Phase-1 Question Tree as a page in the
microsite Reports menu (src/docs/reports/question-tree.{adoc,puml}).

Nodes carry two marks: blue for the 43 fixed nodes the recovery skill
prescribes, orange for the 74 nodes the LLM decomposed from code, and a
red (OPEN) marker on the 14 leaves the code cannot answer that are
handed to the team.

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>

Author: raifdmueller

src/docs/reports/question-tree.adoc

@@ -0,0 +1,35 @@
+:jbake-title: Question Tree
+:jbake-type: page
+:jbake-status: published
+:jbake-menu: reports
+:jbake-order: 0
+ifndef::imagesdir[:imagesdir: ../images]
+
+= Vibe — Question Tree
+
+This is the Question Tree recovered in Phase 1 of Socratic Code-Theory
+Recovery for the `vibe/` bounded context — the reasoning trace that
+separates what the code can answer from what only the team can.
+
+The tree has two kinds of node:
+
+* *Fixed nodes* (blue) are prescribed by the recovery skill — the five
+  root questions Q1–Q5 and the fixed second level Q1.1–Q5.5. They are
+  emitted on every run, so trees from different runs stay diffable.
+* *Adaptive nodes* (orange) are decomposed by the LLM from code
+  evidence. Their depth tracks how much theory each area carries — a
+  dense subsystem yields a deep sub-tree, a thin one a shallow one.
+
+Leaves carry a second mark: a leaf the code can answer is left plain —
+its evidence is a `file:line` reference — while a leaf the code
+*cannot* answer is flagged `(OPEN)` in red and routed to the team.
+
+Of the 117 question nodes, 43 are fixed skeleton and 74 were recovered
+adaptively; 14 leaves are `(OPEN)`. The full tree with `file:line`
+evidence, and the `(OPEN)` leaves grouped by role, live in
+`QUESTION_TREE.adoc` and `OPEN_QUESTIONS.adoc` at the repository root.
+
+[plantuml,question-tree,svg]
+----
+include::question-tree.puml[]
+----

src/docs/reports/question-tree.puml

@@ -0,0 +1,138 @@
+@startmindmap
+title Question Tree — Vibe: skill-fixed vs. LLM-recovered nodes; (OPEN) = handed to the team
+
+' Socratic Code-Theory Recovery — Phase 1 tree visualised.
+' Blue   = fixed node prescribed by the skill (Q1-Q5 and the second level Q1.1-Q5.5).
+' Orange = adaptive node, decomposed by the LLM from code evidence.
+' (OPEN) = leaf the code cannot answer; handed to the team via OPEN_QUESTIONS.adoc.
+
++[#E8E8E8] Question Tree\nVibe
+
+++[#B3D9FF] Q1 — Problem & Users
++++[#B3D9FF] Q1.1 Product identity
++++[#B3D9FF] Q1.2 Primary users
+++++[#FFE0B3] Q1.2.1 Actor roles
+++++[#FFE0B3] Q1.2.2 Segment priority <color:#C62828>(OPEN)</color>
++++[#B3D9FF] Q1.3 Channels
++++[#B3D9FF] Q1.4 Why built <color:#C62828>(OPEN)</color>
++++[#B3D9FF] Q1.5 Success metrics
+++++[#FFE0B3] Q1.5.1 What the code measures
+++++[#FFE0B3] Q1.5.2 Business KPIs <color:#C62828>(OPEN)</color>
++++[#B3D9FF] Q1.6 Segment prioritisation <color:#C62828>(OPEN)</color>
+
+++[#B3D9FF] Q2 — Specification
++++[#B3D9FF] Q2.1 Actors
++++[#B3D9FF] Q2.2 Use-case catalog
+++++[#FFE0B3] PUC InteractiveChat
++++++[#FFE0B3] Actor
++++++[#FFE0B3] Trigger
++++++[#FFE0B3] Main Success
++++++[#FFE0B3] Extensions
++++++[#FFE0B3] Postconditions
+++++[#FFE0B3] PUC ProgrammaticRun
++++++[#FFE0B3] Actor
++++++[#FFE0B3] Main Success
++++++[#FFE0B3] Extensions
+++++[#FFE0B3] PUC ResumeSession
+++++[#FFE0B3] PUC PlanMode
+++++[#FFE0B3] PUC DelegateToSubagent
+++++[#FFE0B3] PUC VoiceMode
+++++[#FFE0B3] PUC Onboard
+++++[#FFE0B3] PUC Teleport
+++++[#FFE0B3] PUC ManageExtensions
+++++[#FFE0B3] UC completeness <color:#C62828>(OPEN)</color>
++++[#B3D9FF] Q2.3 Per-interface system specs
+++++[#FFE0B3] SUC VibeCLI
+++++[#FFE0B3] SUC SlashCommands
+++++[#FFE0B3] SUC AcpServer
+++++[#FFE0B3] SUC ToolInterface
++++++[#FFE0B3] bash
++++++[#FFE0B3] read / write
++++++[#FFE0B3] grep
++++++[#FFE0B3] web
++++++[#FFE0B3] agent tools
+++++[#FFE0B3] SUC ConfigFile
+++++[#FFE0B3] SUC McpInterface
+++++[#FFE0B3] SUC HookInterface
++++[#B3D9FF] Q2.4 Entity model
+++++[#FFE0B3] Session
+++++[#FFE0B3] Conversation
+++++[#FFE0B3] AgentStats
+++++[#FFE0B3] Config
+++++[#FFE0B3] AgentProfile
+++++[#FFE0B3] Checkpoint
++++[#B3D9FF] Q2.5 Acceptance criteria
+++++[#FFE0B3] Q2.5.1 Tests as criteria
+++++[#FFE0B3] Q2.5.2 Gherkin per use case <color:#C62828>(OPEN)</color>
++++[#B3D9FF] Q2.6 Cross-cutting business rules
+++++[#FFE0B3] BR TrustGate
+++++[#FFE0B3] BR ToolApproval
+++++[#FFE0B3] BR AgentSafety
+++++[#FFE0B3] BR TelemetryGate
+++++[#FFE0B3] BR intent <color:#C62828>(OPEN)</color>
+
+--[#B3D9FF] Q3 — Architecture
+---[#B3D9FF] Q3.1 Intro & goals
+---[#B3D9FF] Q3.2 Constraints
+---[#B3D9FF] Q3.3 Context & scope
+---[#B3D9FF] Q3.4 Solution strategy
+----[#FFE0B3] Q3.4.1 Patterns used
+----[#FFE0B3] Q3.4.2 Why chosen <color:#C62828>(OPEN)</color>
+---[#B3D9FF] Q3.5 Building-block view
+----[#FFE0B3] Q3.5.1 Level-1 sub-contexts
+----[#FFE0B3] Q3.5.2 Agent Loop
+----[#FFE0B3] Q3.5.3 LLM Backend Layer
+----[#FFE0B3] Q3.5.4 Tool Subsystem
+----[#FFE0B3] Q3.5.5 Config Subsystem
+----[#FFE0B3] Q3.5.6 Session Persistence
+----[#FFE0B3] Q3.5.7 CLI / Textual UI
+----[#FFE0B3] Q3.5.8 ACP Bridge
+----[#FFE0B3] Q3.5.9 Remote
+---[#B3D9FF] Q3.6 Runtime view
+----[#FFE0B3] SCN Turn
+----[#FFE0B3] SCN Approval
+----[#FFE0B3] SCN Compaction
+----[#FFE0B3] SCN Interrupt
+----[#FFE0B3] SCN RateLimit
+---[#B3D9FF] Q3.7 Deployment view
+---[#B3D9FF] Q3.8 Crosscutting concepts
+----[#FFE0B3] Q3.8.1 Threat model <color:#C62828>(OPEN)</color>
+----[#FFE0B3] Q3.8.2 Security
+----[#FFE0B3] Q3.8.3 Test
+----[#FFE0B3] Q3.8.4 Observability
+----[#FFE0B3] Q3.8.5 Error handling
+----[#FFE0B3] Q3.8.6 Configuration
+---[#B3D9FF] Q3.9 Architecture decisions
+----[#FFE0B3] ADR HexagonalPorts <color:#C62828>(OPEN)</color>
+----[#FFE0B3] ADR LLMBackendAbstraction <color:#C62828>(OPEN)</color>
+----[#FFE0B3] ADR MiddlewarePipeline <color:#C62828>(OPEN)</color>
+----[#FFE0B3] ADR SessionsAsFiles <color:#C62828>(OPEN)</color>
+----[#FFE0B3] ADR index
+---[#B3D9FF] Q3.10 Quality requirements
+---[#B3D9FF] Q3.11 Risks & technical debt
+---[#B3D9FF] Q3.12 Glossary
+
+--[#B3D9FF] Q4 — Quality goals
+---[#B3D9FF] Q4.1 Functional suitability
+---[#B3D9FF] Q4.2 Performance efficiency
+---[#B3D9FF] Q4.3 Compatibility
+---[#B3D9FF] Q4.4 Usability
+---[#B3D9FF] Q4.5 Reliability
+---[#B3D9FF] Q4.6 Security
+---[#B3D9FF] Q4.7 Maintainability
+---[#B3D9FF] Q4.8 Portability
+---[#B3D9FF] Q4.9 Priority ranking <color:#C62828>(OPEN)</color>
+
+--[#B3D9FF] Q5 — Risks & technical debt
+---[#B3D9FF] Q5.1 Technical debt
+---[#B3D9FF] Q5.2 Security risks
+---[#B3D9FF] Q5.3 Operational risks
+---[#B3D9FF] Q5.4 Supply-chain risks
+---[#B3D9FF] Q5.5 Scaling / performance
+
+legend right
+|<#B3D9FF>| Fixed node — prescribed by the Socratic Code-Theory Recovery skill (Q1-Q5, Q1.1-Q5.5) |
+|<#FFE0B3>| Adaptive node — decomposed by the LLM from code evidence |
+| <color:#C62828>(OPEN)</color> | Leaf the code cannot answer — handed to the team; every other leaf is answered from code with file:line evidence |
+endlegend
+@endmindmap