maestro-flow 0.3.21 → 0.3.23

package/.codex/skills/maestro-roadmap/SKILL.md

@@ -1,332 +1,443 @@
----
-name: maestro-roadmap
-description: Lightweight roadmap generation via 2-wave CSV pipeline. Wave 1 runs parallel requirement analysis agents (scope, risk, dependency). Wave 2 runs roadmap assembly agent producing roadmap.md with phases, milestones, and success criteria. Replaces maestro-roadmap command.
-argument-hint: "\"<requirements>\" [-y|--yes] [--phases N] [--from-brainstorm SESSION-ID]"
-allowed-tools: spawn_agents_on_csv, Read, Write, Edit, Bash, Glob, Grep, AskUserQuestion
----
-
-<purpose>
-2-wave roadmap generation using `spawn_agents_on_csv`. Wave 1 runs parallel requirement analysis agents to assess scope, risk, and dependencies. Wave 2 runs a single roadmap assembly agent that synthesizes analysis findings into a complete roadmap with phases, milestones, and success criteria.
-
-**Core workflow**: Parse Requirements -> Parallel Analysis -> Roadmap Assembly -> Interactive Refinement -> Output
-
-```
-+---------------------------------------------------------------------------+
-|                    ROADMAP CSV WAVE WORKFLOW                               |
-+---------------------------------------------------------------------------+
-|                                                                           |
-|  Phase 1: Requirement Parsing + CSV Generation                            |
-|     +-- Parse requirement text from arguments                             |
-|     +-- Detect input type (text, @file, brainstorm import)                |
-|     +-- Codebase detection (conditional exploration)                      |
-|     +-- Assess uncertainty factors for strategy selection                  |
-|     +-- Select decomposition strategy (progressive/direct/auto)           |
-|     +-- Generate tasks.csv with analysis + assembly rows                  |
-|     +-- User validates breakdown (skip if -y)                             |
-|                                                                           |
-|  Phase 2: Wave Execution Engine                                           |
-|     +-- Wave 1: Requirement Analysis (parallel)                           |
-|     |   +-- Scope analysis agent: features, boundaries, MVP definition    |
-|     |   +-- Risk analysis agent: technical risks, uncertainties, mitigations|
-|     |   +-- Dependency analysis agent: ordering, blockers, parallel groups |
-|     |   +-- Discoveries shared via board (constraints, patterns)          |
-|     |   +-- Results: findings per analysis dimension                      |
-|     +-- Wave 2: Roadmap Assembly (sequential)                             |
-|     |   +-- Single agent produces roadmap.md:                             |
-|     |   |   - Phase structure with goals and success criteria             |
-|     |   |   - Dependency ordering between phases                          |
-|     |   |   - Milestone grouping (MVP, Usable, Complete)                  |
-|     |   |   - Scope decisions (in/deferred/out)                           |
-|     |   |   - Progress tracking table                                     |
-|     |   +-- Uses Wave 1 analysis as prev_context                          |
-|     |   +-- Applies decomposition strategy from Phase 1                   |
-|     +-- discoveries.ndjson shared across all waves (append-only)          |
-|                                                                           |
-|  Phase 3: Results Aggregation                                             |
-|     +-- Export results.csv                                                |
-|     +-- Interactive refinement (max 3 rounds, skip if -y)                 |
-|     +-- Generate context.md with roadmap summary                          |
-|     +-- Write .workflow/roadmap.md                                        |
-|     +-- Ensure .workflow/scratch/ directory exists                         |
-|     +-- Display summary with next steps                                   |
-|                                                                           |
-+---------------------------------------------------------------------------+
-```
-</purpose>
-
-<context>
-$ARGUMENTS -- requirement text or @file reference, plus optional flags.
-
-**Usage**:
-
-```bash
-$maestro-roadmap "Implement user authentication with OAuth and 2FA"
-$maestro-roadmap -y "@requirements.md"
-$maestro-roadmap --phases 4 "Build real-time notification system"
-$maestro-roadmap --from-brainstorm WFS-001 "Enhance auth system"
-```
-
-**Flags**:
-- `-y, --yes`: Skip all confirmations (auto mode)
-- `--phases N`: Target number of roadmap phases (default: auto-determined)
-- `--from-brainstorm SESSION-ID`: Import guidance-specification.md from brainstorm session as seed
-
-When `--yes` or `-y`: Auto-confirm strategy selection, skip interactive refinement rounds, use recommended defaults for decomposition mode and phase count.
-
-**Output Directory**: `.workflow/.csv-wave/{session-id}/`
-**Core Output**: `tasks.csv` (master state) + `results.csv` (final) + `discoveries.ndjson` (shared exploration) + `context.md` (human-readable report) + `.workflow/roadmap.md`
-</context>
-
-<csv_schema>
-
-### tasks.csv (Master State)
-
-```csv
-id,title,description,analysis_focus,deps,context_from,wave,status,findings,error
-"1","Scope Analysis","Analyze requirement scope: identify all features and sub-features, define MVP boundaries, classify must-have vs nice-to-have, estimate relative size of each feature area. Produce feature inventory with priority tags.","scope","","","1","","",""
-"2","Risk Analysis","Assess technical and project risks: identify unknowns, evaluate technical feasibility per feature, rate risk levels (high/medium/low), propose mitigations. Flag features requiring spikes or prototypes.","risk","","","1","","",""
-"3","Dependency Analysis","Map dependencies between features: identify ordering constraints, find parallel-safe groups, detect external dependencies (APIs, libraries, infrastructure). Produce dependency graph with critical path.","dependency","","","1","","",""
-"4","Roadmap Assembly","Synthesize analysis findings into a complete roadmap. Apply decomposition strategy. Produce roadmap.md with: phase structure (goal, depends-on, requirements, success criteria), milestone grouping, scope decisions, progress table. Each phase must have observable success criteria from user perspective.","","1;2;3","1;2;3","2","","",""
-```
-
-**Columns**:
-
-| Column | Phase | Description |
-|--------|-------|-------------|
-| `id` | Input | Unique task identifier (string) |
-| `title` | Input | Short task title |
-| `description` | Input | Detailed instructions for this task |
-| `analysis_focus` | Input | Analysis dimension: scope/risk/dependency (Wave 1 only) |
-| `deps` | Input | Semicolon-separated dependency task IDs |
-| `context_from` | Input | Semicolon-separated task IDs whose findings this task needs |
-| `wave` | Computed | Wave number (1 = analysis, 2 = assembly) |
-| `status` | Output | `pending` -> `completed` / `failed` / `skipped` |
-| `findings` | Output | Key findings summary (max 500 chars) |
-| `error` | Output | Error message if failed |
-
-### Per-Wave CSV (Temporary)
-
-Each wave generates `wave-{N}.csv` with extra `prev_context` column.
-</csv_schema>
-
-<invariants>
-1. **Start Immediately**: First action is session initialization, then Phase 1
-2. **Wave Order is Sacred**: Never execute wave 2 before wave 1 completes and results are merged
-3. **CSV is Source of Truth**: Master tasks.csv holds all state
-4. **Context Propagation**: prev_context built from master CSV, not from memory
-5. **Discovery Board is Append-Only**: Never clear, modify, or recreate discoveries.ndjson
-6. **Graceful Degradation**: If analysis fails, assembly proceeds with requirement text only
-7. **Cleanup Temp Files**: Remove wave-{N}.csv after results are merged
-8. **DO NOT STOP**: Continuous execution until all waves complete
-</invariants>
-
-<execution>
-
-### Output Artifacts
-
-| File | Purpose | Lifecycle |
-|------|---------|-----------|
-| `tasks.csv` | Master state -- all tasks with status/findings | Updated after each wave |
-| `wave-{N}.csv` | Per-wave input (temporary) | Created before wave, deleted after |
-| `results.csv` | Final export of all task results | Created in Phase 3 |
-| `discoveries.ndjson` | Shared exploration board | Append-only, carries across waves |
-| `context.md` | Human-readable roadmap generation report | Created in Phase 3 |
-
-### Session Structure
-
-```
-.workflow/.csv-wave/{YYYYMMDD}-roadmap-{slug}/
-+-- tasks.csv
-+-- results.csv
-+-- discoveries.ndjson
-+-- context.md
-+-- wave-{N}.csv (temporary)
-```
-
-### Session Initialization
-
-Parse `$ARGUMENTS` to extract:
-- `AUTO_YES` from `--yes` / `-y`
-- `targetPhases` from `--phases N` (default: auto-determined)
-- `brainstormSession` from `--from-brainstorm <SESSION-ID>`
-- `requirementArg` = remaining text after stripping all flags
-- `slug` = requirementArg lowercased, non-alphanumeric → `-`, max 40 chars
-
-Session ID: `{YYYYMMDD}-roadmap-{slug}`
-Session folder: `.workflow/.csv-wave/{sessionId}/` — create via `mkdir -p`
-
-### Phase 1: Requirement Parsing + CSV Generation
-
-**Objective**: Parse requirements, assess uncertainty, select strategy, generate tasks.csv.
-
-**Decomposition Rules**:
-
-1. **Input parsing**: Parse `{requirementArg}` -- direct text or `@file` reference
-2. **Brainstorm import**: If `--from-brainstorm`, read `guidance-specification.md` for enriched context (problem statement, features, non-goals, terminology)
-3. **Codebase detection**: Check for source files; if found, add codebase context to analysis prompts
-4. **Load project specs**: Read `.workflow/specs/` for constraint awareness
-
-5. **Uncertainty assessment**:
-
-| Factor | Low | Medium | High |
-|--------|-----|--------|------|
-| Scope clarity | Requirements explicit | Some ambiguity | Vague/open-ended |
-| Technical risk | Proven stack | Some unknowns | New technology |
-| Dependency unknown | All mapped | Some unclear | Many external |
-| Domain familiarity | Expert | Moderate | New domain |
-| Requirement stability | Locked | Some flux | Evolving |
-
-Strategy: >= 3 high -> progressive, >= 3 low -> direct, else -> ask user (or auto if `-y`).
-
-6. **CSV generation**: 3 analysis tasks (wave 1) + 1 assembly task (wave 2).
-
-7. **Wave computation**: Simple 2-wave -- all analysis tasks = wave 1, assembly = wave 2.
-
-**User validation**: Display task breakdown + strategy (skip if AUTO_YES).
-
-### Phase 2: Wave Execution Engine
-
-**Objective**: Execute analysis and assembly wave-by-wave via spawn_agents_on_csv.
-
-#### Wave 1: Requirement Analysis (Parallel)
-
-Filter master `tasks.csv` for `wave == 1 AND status == pending` → write `wave-1.csv` (no prev_context needed).
-
-```javascript
-spawn_agents_on_csv({
-  csv_path: `${sessionFolder}/wave-1.csv`,
-  id_column: "id",
-  instruction: buildAnalysisInstruction(sessionFolder, requirementArg, strategy),
-  max_concurrency: 3,
-  max_runtime_seconds: 3600,
-  output_csv_path: `${sessionFolder}/wave-1-results.csv`,
-  output_schema: {
-    type: "object",
-    properties: {
-      id: { type: "string" },
-      status: { type: "string", enum: ["completed", "failed"] },
-      findings: { type: "string" },
-      error: { type: "string" }
-    },
-    required: ["id", "status", "findings"]
-  }
-})
-```
-
-Merge `wave-1-results.csv` into master `tasks.csv`, delete `wave-1.csv`.
-
-#### Wave 2: Roadmap Assembly (Sequential)
-
-Filter master `tasks.csv` for `wave == 2 AND status == pending`. If all wave 1 tasks failed, use degraded mode (requirement text only).
-
-Build `prev_context` from wave 1 findings (format: `[Task N: Title] summary...` per task). Inject strategy and `--phases N` constraint into assembly prompt.
-Write `wave-2.csv` with `prev_context` column → execute `spawn_agents_on_csv` → merge results → delete `wave-2.csv`.
-
-### Phase 3: Results Aggregation
-
-**Objective**: Generate final results, optional refinement, and write roadmap.
-
-Export master `tasks.csv` as `results.csv`.
-
-**Interactive refinement** (skip if AUTO_YES): Present roadmap overview, collect user feedback via AskUserQuestion (max 3 rounds). Options: Approve, Adjust Scope, Reorder, Split/Merge. Update draft each round.
-
-Generate `context.md`:
-
-```markdown
-# Roadmap Generation Report
-
-## Summary
-- Requirements: {requirement_summary}
-- Strategy: {progressive|direct}
-- Analysis agents: 3 ({completed_count} completed)
-- Phases generated: {phase_count}
-- Milestones: {milestone_count}
-
-## Analysis Findings
-### Scope Analysis
-{findings}
-
-### Risk Analysis
-{findings}
-
-### Dependency Analysis
-{findings}
-
-## Roadmap
-- Phases: {phase_count}
-- Strategy: {strategy}
-- MVP scope: {mvp_description}
-- Deferred: {deferred_items}
-```
-
-**Write outputs**: `.workflow/roadmap.md` (standard template), ensure `.workflow/scratch/` exists, update `state.json` milestones array and `current_milestone`.
-
-Display summary:
-
-```
-=== ROADMAP CREATED ===
-Strategy: {progressive|direct}
-Phases:   {phase_count} across {milestone_count} milestones
-Roadmap:  .workflow/roadmap.md
-
-Next steps:
-  maestro-init                    -- Set up project (if not yet initialized)
-  maestro-plan "1"                -- Plan first phase
-  manage-status                   -- View project dashboard
-```
-
-### Shared Discovery Board Protocol
-
-#### Standard Discovery Types
-
-| Type | Dedup Key | Data Schema | Description |
-|------|-----------|-------------|-------------|
-| `code_pattern` | `data.name` | `{name, file, description}` | Reusable code pattern found |
-| `integration_point` | `data.file` | `{file, description, exports[]}` | Module connection point |
-| `convention` | singleton | `{naming, imports, formatting}` | Project code conventions |
-| `tech_stack` | singleton | `{framework, language, tools[]}` | Technology stack info |
-
-#### Domain Discovery Types
-
-| Type | Dedup Key | Data Schema | Description |
-|------|-----------|-------------|-------------|
-| `scope_boundary` | `data.feature` | `{feature, inclusion, rationale}` | Scope inclusion/exclusion decision |
-| `risk_factor` | `data.name` | `{name, severity, probability, mitigation}` | Identified risk |
-| `dependency_constraint` | `data.from+data.to` | `{from, to, type, strength}` | Dependency between features |
-| `external_dependency` | `data.name` | `{name, type, risk, alternative}` | External system dependency |
-
-#### Protocol
-
-Read `{session_folder}/discoveries.ndjson` before own analysis. Deduplicate by type + dedup key before writing. Append-only — never modify or delete.
-
-```bash
-echo '{"ts":"<ISO>","worker":"{id}","type":"risk_factor","data":{"name":"OAuth provider rate limits","severity":"medium","probability":"high","mitigation":"Implement token caching and retry logic"}}' >> {session_folder}/discoveries.ndjson
-```
-</execution>
-
-<error_codes>
-| Error | Resolution |
-|-------|------------|
-| No requirement text provided | Abort with error: "Requirement text or @file required" |
-| Brainstorm session not found | Abort with error: "Session {id} not found" -- list available sessions |
-| @file not found | Abort with error: "File {path} not found" |
-| Analysis agent timeout | Mark as failed, assembly uses available findings |
-| All analysis agents failed | Assembly runs in degraded mode (requirement text only) |
-| Assembly agent failed | Abort with error: "Roadmap generation failed" |
-| Circular dependency detected | Prompt user to re-decompose |
-| CSV parse error | Validate format, show line number |
-| discoveries.ndjson corrupt | Ignore malformed lines |
-| Max refinement rounds (3) | Force proceed with current roadmap |
-</error_codes>
-
-<success_criteria>
-- [ ] Requirements parsed from text, @file, or brainstorm session
-- [ ] Uncertainty assessed and decomposition strategy selected
-- [ ] Wave 1 analysis agents completed (scope, risk, dependency)
-- [ ] Wave 2 assembly agent produced roadmap.md
-- [ ] Interactive refinement offered (or skipped with -y)
-- [ ] .workflow/roadmap.md written with phases, milestones, success criteria
-- [ ] state.json updated with milestones
-- [ ] context.md generated with analysis summary
-- [ ] Completion report displayed with next steps
-</success_criteria>
+---
+name: maestro-roadmap
+description: Dual-mode roadmap generation via 2-wave CSV pipeline. Light mode (default) runs scope/risk/dependency analysis → roadmap assembly. Full mode (--mode full) runs domain/competitive/tech-stack research → 7-phase specification document chain → roadmap. Both modes produce .workflow/roadmap.md.
+argument-hint: "\"<requirements>\" [--mode light|full] [-y|--yes] [--phases N] [--skip-research] [--from-brainstorm SESSION-ID]"
+allowed-tools: spawn_agents_on_csv, Read, Write, Edit, Bash, Glob, Grep, AskUserQuestion
+---
+
+<purpose>
+Unified 2-wave roadmap generation using `spawn_agents_on_csv` with dual execution modes:
+
+- **Light mode** (default): Wave 1 runs parallel requirement analysis agents (scope, risk, dependency). Wave 2 runs roadmap assembly agent producing `roadmap.md` with phases, milestones, and success criteria.
+- **Full mode** (`--mode full`): Wave 1 runs parallel research agents (domain, competitive, tech stack). Wave 2 runs a single synthesis agent producing the 7-phase specification document chain (product brief, PRD, architecture, epics, roadmap).
+
+**Common workflow**: Parse Input → Parallel Analysis/Research → Sequential Assembly/Synthesis → Output
+
+```
++---------------------------------------------------------------------------+
+|               ROADMAP CSV WAVE WORKFLOW (DUAL MODE)                       |
++---------------------------------------------------------------------------+
+|                                                                           |
+|  Phase 1: Input Parsing + CSV Generation                                  |
+|     +-- Parse requirement/idea from arguments                             |
+|     +-- Detect input type (text, @file, brainstorm import)                |
+|     +-- Detect mode: light (default) or full (--mode full)                |
+|     +-- Codebase detection (conditional exploration)                      |
+|     +-- Generate mode-specific tasks.csv                                  |
+|     +-- User validates breakdown (skip if -y)                             |
+|                                                                           |
+|  Phase 2: Wave Execution Engine                                           |
+|     +-- Wave 1: Analysis/Research (parallel)                              |
+|     |   [Light] scope, risk, dependency analysis                          |
+|     |   [Full]  domain, competitive, tech stack research                  |
+|     +-- Wave 2: Assembly/Synthesis (sequential)                           |
+|     |   [Light] roadmap assembly agent → roadmap.md                       |
+|     |   [Full]  7-phase document chain → spec package + roadmap.md        |
+|     +-- discoveries.ndjson shared across waves (append-only)              |
+|                                                                           |
+|  Phase 3: Results Aggregation                                             |
+|     +-- Export results.csv                                                |
+|     +-- Interactive refinement (skip if -y)                               |
+|     +-- [Full only] Readiness check (4 dimensions)                        |
+|     +-- Generate context.md                                               |
+|     +-- Write .workflow/roadmap.md                                        |
+|     +-- Display summary with next steps                                   |
+|                                                                           |
++---------------------------------------------------------------------------+
+```
+</purpose>
+
+<context>
+$ARGUMENTS -- requirement/idea text or @file reference, plus optional flags.
+
+**Usage**:
+
+```bash
+# Light mode (default)
+$maestro-roadmap "Implement user authentication with OAuth and 2FA"
+$maestro-roadmap -y "@requirements.md"
+$maestro-roadmap --phases 4 "Build real-time notification system"
+
+# Full mode
+$maestro-roadmap --mode full "Build a real-time collaboration platform"
+$maestro-roadmap --mode full -y "@requirements.md"
+$maestro-roadmap --mode full --skip-research "CLI workflow orchestration tool"
+
+# Shared flags
+$maestro-roadmap --from-brainstorm WFS-001 "Enhance auth system"
+```
+
+**Flags**:
+- `--mode light|full`: Execution mode (default: light)
+- `-y, --yes`: Skip all confirmations (auto mode)
+- `--phases N`: Target number of roadmap phases (default: auto-determined, light mode only)
+- `--skip-research`: Skip Wave 1 research, jump to document generation (full mode only)
+- `--from-brainstorm SESSION-ID`: Import guidance-specification.md from brainstorm session as seed
+
+When `--yes` or `-y`: Auto-confirm strategy/decisions, skip interactive refinement, use defaults.
+
+**Output Directory**: `.workflow/.csv-wave/{session-id}/`
+**Core Output**: `tasks.csv` + `results.csv` + `discoveries.ndjson` + `context.md` + `.workflow/roadmap.md`
+**Full mode additional**: Spec package in `.workflow/.spec/SPEC-{slug}-{date}/`
+</context>
+
+<csv_schema>
+
+### tasks.csv — Light Mode
+
+```csv
+id,title,description,analysis_focus,deps,context_from,wave,status,findings,error
+"1","Scope Analysis","Analyze requirement scope: identify all features and sub-features, define MVP boundaries, classify must-have vs nice-to-have, estimate relative size of each feature area. Produce feature inventory with priority tags.","scope","","","1","","",""
+"2","Risk Analysis","Assess technical and project risks: identify unknowns, evaluate technical feasibility per feature, rate risk levels (high/medium/low), propose mitigations. Flag features requiring spikes or prototypes.","risk","","","1","","",""
+"3","Dependency Analysis","Map dependencies between features: identify ordering constraints, find parallel-safe groups, detect external dependencies (APIs, libraries, infrastructure). Produce dependency graph with critical path.","dependency","","","1","","",""
+"4","Roadmap Assembly","Synthesize analysis findings into a complete roadmap. Apply decomposition strategy. Produce roadmap.md with: phase structure (goal, depends-on, requirements, success criteria), milestone grouping, scope decisions, progress table. Each phase must have observable success criteria from user perspective.","","1;2;3","1;2;3","2","","",""
+```
+
+### tasks.csv — Full Mode
+
+```csv
+id,title,description,research_focus,doc_phase,deps,context_from,wave,status,findings,output_file,error
+"1","Domain Research","Research the problem domain: identify target users, market needs, existing solutions, industry trends, and domain terminology. Produce structured findings with confidence levels.","domain","","","","1","","","",""
+"2","Competitive Analysis","Analyze competing products and approaches: feature comparison matrix, UX patterns, pricing models, market positioning. Identify gaps and opportunities for differentiation.","competitive","","","","1","","","",""
+"3","Tech Stack Analysis","Evaluate technical feasibility: recommended languages, frameworks, databases, infrastructure. Assess constraints, integration complexity, scalability requirements. Reference existing codebase if available.","tech_stack","","","","1","","","",""
+"4","Document Chain","Generate complete 7-phase specification package using research context. Phases: (1) Product Brief, (2) PRD with REQ-*/NFR-*, (3) Architecture with ADR-*, (4) Data Model, (5) API Specification, (6) UI Wireframes, (7) Epic-to-Roadmap with EPIC-* and phase mapping. Produce glossary.json for terminology consistency.","","1-7","1;2;3","1;2;3","2","","","",""
+```
+
+### Shared Columns
+
+| Column | Phase | Description |
+|--------|-------|-------------|
+| `id` | Input | Unique task identifier (string) |
+| `title` | Input | Short task title |
+| `description` | Input | Detailed instructions for this task |
+| `deps` | Input | Semicolon-separated dependency task IDs |
+| `context_from` | Input | Semicolon-separated task IDs whose findings this task needs |
+| `wave` | Computed | Wave number (1 = analysis/research, 2 = assembly/synthesis) |
+| `status` | Output | `pending` -> `completed` / `failed` / `skipped` |
+| `findings` | Output | Key findings summary (max 500 chars) |
+| `error` | Output | Error message if failed |
+
+**Light-only**: `analysis_focus` (scope/risk/dependency)
+**Full-only**: `research_focus` (domain/competitive/tech_stack), `doc_phase`, `output_file`
+
+### Per-Wave CSV (Temporary)
+
+Each wave generates `wave-{N}.csv` with extra `prev_context` column.
+</csv_schema>
+
+<invariants>
+1. **Start Immediately**: First action is session initialization, then Phase 1
+2. **Wave Order is Sacred**: Never execute wave 2 before wave 1 completes and results are merged
+3. **CSV is Source of Truth**: Master tasks.csv holds all state
+4. **Context Propagation**: prev_context built from master CSV, not from memory
+5. **Discovery Board is Append-Only**: Never clear, modify, or recreate discoveries.ndjson
+6. **Graceful Degradation**: If Wave 1 fails, Wave 2 proceeds with seed input only
+7. **Cleanup Temp Files**: Remove wave-{N}.csv after results are merged
+8. **DO NOT STOP**: Continuous execution until all waves complete
+</invariants>
+
+<execution>
+
+### Output Artifacts
+
+| File | Purpose | Lifecycle |
+|------|---------|-----------|
+| `tasks.csv` | Master state -- all tasks with status/findings | Updated after each wave |
+| `wave-{N}.csv` | Per-wave input (temporary) | Created before wave, deleted after |
+| `results.csv` | Final export of all task results | Created in Phase 3 |
+| `discoveries.ndjson` | Shared exploration board | Append-only, carries across waves |
+| `context.md` | Human-readable report | Created in Phase 3 |
+| `spec-config.json` | Session metadata (full mode only) | Created in Phase 1 |
+
+### Session Structure
+
+**Light mode**:
+```
+.workflow/.csv-wave/{YYYYMMDD}-roadmap-{slug}/
++-- tasks.csv, results.csv, discoveries.ndjson, context.md
+```
+
+**Full mode**:
+```
+.workflow/.csv-wave/{YYYYMMDD}-roadmap-full-{slug}/
++-- tasks.csv, results.csv, discoveries.ndjson, context.md, spec-config.json
+
+.workflow/.spec/SPEC-{slug}-{date}/
++-- spec-config.json
++-- product-brief.md
++-- glossary.json
++-- requirements/
+|   +-- _index.md
+|   +-- REQ-NNN-{slug}.md
+|   +-- NFR-{type}-NNN-{slug}.md
++-- architecture/
+|   +-- _index.md
+|   +-- ADR-NNN-{slug}.md
++-- epics/
+|   +-- _index.md
+|   +-- EPIC-NNN-{slug}.md
++-- readiness-report.md
++-- spec-summary.md
++-- roadmap.md
+```
+
+### Session Initialization
+
+Parse `$ARGUMENTS` to extract:
+- `mode` from `--mode light|full` (default: `light`)
+- `AUTO_YES` from `--yes` / `-y`
+- `targetPhases` from `--phases N` (light mode only)
+- `skipResearch` from `--skip-research` (full mode only)
+- `brainstormSession` from `--from-brainstorm <SESSION-ID>`
+- `requirementArg` = remaining text after stripping all flags
+- `slug` = requirementArg lowercased, non-alphanumeric → `-`, max 40 chars
+
+Session ID: `{YYYYMMDD}-roadmap-{slug}` (light) or `{YYYYMMDD}-roadmap-full-{slug}` (full)
+Session folder: `.workflow/.csv-wave/{sessionId}/` — create via `mkdir -p`
+
+### Phase 1: Input Parsing + CSV Generation
+
+**Shared steps**:
+1. **Input parsing**: Parse `{requirementArg}` -- direct text or `@file` reference
+2. **Brainstorm import**: If `--from-brainstorm`, read `guidance-specification.md` for enriched context
+3. **Codebase detection**: Check for source files; if found, add context to analysis/research prompts
+4. **Load project specs**: Read `.workflow/specs/` for constraint awareness
+
+**Light mode specific**:
+5. **Uncertainty assessment**:
+
+| Factor | Low | Medium | High |
+|--------|-----|--------|------|
+| Scope clarity | Requirements explicit | Some ambiguity | Vague/open-ended |
+| Technical risk | Proven stack | Some unknowns | New technology |
+| Dependency unknown | All mapped | Some unclear | Many external |
+| Domain familiarity | Expert | Moderate | New domain |
+| Requirement stability | Locked | Some flux | Evolving |
+
+Strategy: >=3 high → progressive, >=3 low → direct, else → ask (or auto if -y).
+
+6. **CSV generation**: 3 analysis tasks (wave 1) + 1 assembly task (wave 2)
+
+**Full mode specific**:
+5. **Session init**: Create `spec-config.json` with session metadata
+6. **CSV generation**: 3 research tasks (wave 1) + 1 document chain task (wave 2). If `--skip-research`: 0 research + 1 doc chain (wave 1).
+
+**User validation**: Display task breakdown + strategy (skip if AUTO_YES).
+
+### Phase 2: Wave Execution Engine
+
+#### Wave 1 (Parallel)
+
+Filter master `tasks.csv` for `wave == 1 AND status == pending` → write `wave-1.csv`.
+
+```javascript
+spawn_agents_on_csv({
+  csv_path: `${sessionFolder}/wave-1.csv`,
+  id_column: "id",
+  instruction: mode === 'light'
+    ? buildAnalysisInstruction(sessionFolder, requirementArg, strategy)
+    : buildResearchInstruction(sessionFolder, requirementArg),
+  max_concurrency: 3,
+  max_runtime_seconds: 3600,
+  output_csv_path: `${sessionFolder}/wave-1-results.csv`,
+  output_schema: {
+    type: "object",
+    properties: {
+      id: { type: "string" },
+      status: { type: "string", enum: ["completed", "failed"] },
+      findings: { type: "string" },
+      output_file: { type: "string" },  // full mode only
+      error: { type: "string" }
+    },
+    required: ["id", "status", "findings"]
+  }
+})
+```
+
+Merge results into master `tasks.csv`, delete `wave-1.csv`.
+
+#### Wave 2 (Sequential)
+
+Filter for `wave == 2 AND status == pending`. If all wave 1 tasks failed, use degraded mode (requirement/seed text only).
+
+Build `prev_context` from wave 1 findings. Format: `[Task N: Title] summary...` per completed task. Inject decomposition strategy and `--phases N` constraint (light mode) into assembly prompt.
+
+Write `wave-2.csv` with `prev_context` column → execute `spawn_agents_on_csv` → merge → delete temp CSV.
+
+**Light mode**: Assembly agent produces roadmap.md with phases, milestones, success criteria.
+**Full mode**: Document chain agent produces 7-phase spec package + glossary.json. If agent fails, export partial output and log quality issues (do NOT abort).
+
+### Phase 3: Results Aggregation
+
+Export master `tasks.csv` as `results.csv`.
+
+**Interactive refinement** (skip if AUTO_YES): Present overview, collect feedback via AskUserQuestion (max 3 rounds).
+
+**Full mode only — Readiness check**: Score on 4 dimensions (25% each):
+- Completeness, Consistency, Traceability, Depth
+- Gate: Pass (>=80%) / Review (60-79%) / Fail (<60%)
+
+Generate `context.md`:
+
+**Light mode template**:
+```markdown
+# Roadmap Generation Report
+
+## Summary
+- Requirements: {requirement_summary}
+- Strategy: {progressive|direct}
+- Analysis agents: 3 ({completed_count} completed)
+- Phases generated: {phase_count}
+- Milestones: {milestone_count}
+
+## Analysis Findings
+### Scope Analysis
+{findings}
+### Risk Analysis
+{findings}
+### Dependency Analysis
+{findings}
+
+## Roadmap
+- Phases: {phase_count}
+- Strategy: {strategy}
+- MVP scope: {mvp_description}
+- Deferred: {deferred_items}
+```
+
+**Full mode template**:
+```markdown
+# Spec Generate Report
+
+## Summary
+- Topic: {topic}
+- Research agents: 3 ({completed_count} completed)
+- Document phases: 7
+- Quality score: {score}% ({gate})
+
+## Research Findings
+### Domain Research
+{findings}
+### Competitive Analysis
+{findings}
+### Tech Stack Analysis
+{findings}
+
+## Document Chain Output
+- Product Brief: {status}
+- Requirements: {req_count} REQs + {nfr_count} NFRs
+- Architecture: {adr_count} ADRs
+- Epics: {epic_count} Epics
+- Roadmap: {phase_count} phases
+
+## Readiness
+- Completeness: {score}%
+- Consistency: {score}%
+- Traceability: {score}%
+- Depth: {score}%
+- Overall: {score}% ({gate})
+```
+
+**Write outputs**:
+- `.workflow/roadmap.md` (both modes)
+- `.workflow/.spec/SPEC-{slug}-{date}/` (full mode only)
+- Ensure `.workflow/scratch/` exists
+- Update `state.json` milestones array and `current_milestone`
+
+Display summary with next steps:
+```
+=== ROADMAP CREATED ===
+Mode:     {light|full}
+Strategy: {progressive|direct}
+Phases:   {phase_count} across {milestone_count} milestones
+Roadmap:  .workflow/roadmap.md
+[Full]    Spec:  .workflow/.spec/SPEC-{slug}-{date}/
+[Full]    Quality: {score}% ({gate})
+
+Next steps:
+  maestro-init                    -- Set up project (if not yet initialized)
+  maestro-plan "1"                -- Plan first phase
+  manage-status                   -- View project dashboard
+```
+
+### Shared Discovery Board Protocol
+
+#### Standard Discovery Types
+
+| Type | Dedup Key | Data Schema | Description |
+|------|-----------|-------------|-------------|
+| `code_pattern` | `data.name` | `{name, file, description}` | Reusable code pattern found |
+| `integration_point` | `data.file` | `{file, description, exports[]}` | Module connection point |
+| `convention` | singleton | `{naming, imports, formatting}` | Project code conventions |
+| `tech_stack` | singleton | `{framework, language, tools[]}` | Technology stack info |
+
+#### Light Mode Discovery Types
+
+| Type | Dedup Key | Data Schema | Description |
+|------|-----------|-------------|-------------|
+| `scope_boundary` | `data.feature` | `{feature, inclusion, rationale}` | Scope decision |
+| `risk_factor` | `data.name` | `{name, severity, probability, mitigation}` | Identified risk |
+| `dependency_constraint` | `data.from+data.to` | `{from, to, type, strength}` | Feature dependency |
+| `external_dependency` | `data.name` | `{name, type, risk, alternative}` | External system dependency |
+
+#### Full Mode Discovery Types
+
+| Type | Dedup Key | Data Schema | Description |
+|------|-----------|-------------|-------------|
+| `domain_term` | `data.term` | `{term, definition, aliases}` | Domain terminology |
+| `competitor` | `data.name` | `{name, features[], gaps[]}` | Competitive product |
+| `market_trend` | `data.name` | `{name, impact, relevance}` | Market trend |
+| `tech_constraint` | `data.name` | `{name, type, severity, mitigation}` | Technical constraint |
+
+#### Protocol
+
+Read `{session_folder}/discoveries.ndjson` before own analysis. Deduplicate by type + dedup key before writing. Append-only — never modify or delete.
+
+```bash
+# Light mode example
+echo '{"ts":"<ISO>","worker":"{id}","type":"risk_factor","data":{"name":"OAuth provider rate limits","severity":"medium","probability":"high","mitigation":"Implement token caching and retry logic"}}' >> {session_folder}/discoveries.ndjson
+
+# Full mode example
+echo '{"ts":"<ISO>","worker":"{id}","type":"domain_term","data":{"term":"workflow","definition":"A sequence of orchestrated tasks","aliases":["pipeline","process"]}}' >> {session_folder}/discoveries.ndjson
+```
+</execution>
+
+<error_codes>
+| Error | Resolution |
+|-------|------------|
+| No requirement/idea text provided | Abort with error: "Requirement text or @file required" |
+| Brainstorm session not found | Abort with error: "Session {id} not found" -- list available sessions |
+| @file not found | Abort with error: "File {path} not found" |
+| Wave 1 agent timeout | Mark as failed, Wave 2 uses available findings |
+| All Wave 1 agents failed | Wave 2 runs in degraded mode (seed input only) |
+| Wave 2 agent failed (light) | Abort with error: "Roadmap generation failed" |
+| Wave 2 agent failed (full) | Export partial output, log quality issues |
+| Circular dependency detected | Prompt user to re-decompose (light mode) |
+| CSV parse error | Validate format, show line number |
+| discoveries.ndjson corrupt | Ignore malformed lines |
+| Max refinement rounds (3) | Force proceed with current output |
+| Readiness score < 60% (full) | Log issues, proceed with available output |
+</error_codes>
+
+<success_criteria>
+
+**Both modes**:
+- [ ] Input parsed from text, @file, or brainstorm session
+- [ ] Wave 1 agents completed (analysis or research)
+- [ ] Wave 2 agent produced output
+- [ ] Interactive refinement offered (or skipped with -y)
+- [ ] .workflow/roadmap.md written
+- [ ] state.json updated with milestones
+- [ ] context.md generated
+- [ ] Completion report displayed with next steps
+
+**Light mode additional**:
+- [ ] Uncertainty assessed and decomposition strategy selected
+- [ ] Roadmap has phases with milestones and success criteria
+
+**Full mode additional**:
+- [ ] spec-config.json created with session metadata
+- [ ] Spec package written to .workflow/.spec/SPEC-{slug}-{date}/
+- [ ] Readiness check scored on 4 dimensions
+- [ ] Quality score and gate reported
+</success_criteria>