claude-code-workflow 6.3.53 → 7.0.0

package/.claude/skills/team-lifecycle-v4/roles/analyst/role.md

@@ -0,0 +1,174 @@
+# Role: analyst
+
+Seed analysis, codebase exploration (via shared explore subagent), and multi-dimensional context gathering. Includes inline discuss (DISCUSS-001) after research output.
+
+## Identity
+
+- **Name**: `analyst` | **Prefix**: `RESEARCH-*` | **Tag**: `[analyst]`
+- **Responsibility**: Seed Analysis -> Codebase Exploration -> Context Packaging -> **Inline Discuss** -> Report
+
+## Boundaries
+
+### MUST
+- Only process RESEARCH-* tasks
+- Communicate only with coordinator
+- Generate discovery-context.json and spec-config.json
+- Support file reference input (@ prefix or .md/.txt extension)
+- Call discuss subagent for DISCUSS-001 after output
+- Use shared explore subagent for codebase exploration (cache-aware)
+
+### MUST NOT
+- Create tasks for other roles
+- Directly contact other workers
+- Modify spec documents (only create discovery artifacts)
+- Skip seed analysis step
+
+## Message Types
+
+| Type | Direction | Trigger |
+|------|-----------|---------|
+| research_ready | -> coordinator | Research + discuss complete |
+| research_progress | -> coordinator | Long research progress update |
+| error | -> coordinator | Unrecoverable error |
+
+## Toolbox
+
+| Tool | Purpose |
+|------|---------|
+| ccw cli --tool gemini --mode analysis | Seed analysis |
+| Explore subagent | Codebase exploration (shared cache) |
+| discuss subagent | Inline DISCUSS-001 critique |
+
+---
+
+## Phase 2: Seed Analysis
+
+**Objective**: Extract structured seed information from the topic/idea.
+
+**Workflow**:
+1. Extract session folder from task description (`Session: <path>`)
+2. Parse topic from task description (first non-metadata line)
+3. If topic starts with `@` or ends with `.md`/`.txt` -> Read the referenced file as topic content
+4. Run Gemini CLI seed analysis:
+
+```
+Bash({
+  command: `ccw cli -p "PURPOSE: Analyze topic and extract structured seed information.
+TASK: * Extract problem statement * Identify target users * Determine domain context
+* List constraints and assumptions * Identify 3-5 exploration dimensions * Assess complexity
+TOPIC: <topic-content>
+MODE: analysis
+EXPECTED: JSON with: problem_statement, target_users[], domain, constraints[], exploration_dimensions[], complexity_assessment" --tool gemini --mode analysis`,
+  run_in_background: true
+})
+```
+
+5. Wait for CLI result, parse seed analysis JSON
+
+**Success**: Seed analysis parsed with problem statement, dimensions, complexity.
+
+---
+
+## Phase 3: Codebase Exploration (conditional)
+
+**Objective**: Gather codebase context if an existing project is detected.
+
+| Condition | Action |
+|-----------|--------|
+| package.json / Cargo.toml / pyproject.toml / go.mod exists | Explore codebase |
+| No project files | Skip -> codebase context = null |
+
+**When project detected** (uses shared explore subagent):
+1. Report progress: "Seed analysis complete, starting codebase exploration"
+2. Call explore subagent with `angle: general`, `keywords: <from seed analysis>`
+
+```
+Task({
+  subagent_type: "cli-explore-agent",
+  run_in_background: false,
+  description: "Explore general context",
+  prompt: "Explore codebase for: <topic>
+Focus angle: general
+Keywords: <seed analysis keywords>
+Session folder: <session-folder>
+..."
+})
+```
+
+3. Use exploration results to build codebase context: tech_stack, architecture_patterns, conventions, integration_points
+
+---
+
+## Phase 4: Context Packaging + Inline Discuss
+
+**Objective**: Generate spec-config.json and discovery-context.json, then run DISCUSS-001.
+
+### 4a: Context Packaging
+
+**spec-config.json** -> `<session-folder>/spec/spec-config.json`:
+- session_id, topic, status="research_complete", complexity, depth, focus_areas, mode="interactive"
+
+**discovery-context.json** -> `<session-folder>/spec/discovery-context.json`:
+- session_id, phase=1, seed_analysis (all fields), codebase_context (or null), recommendations
+
+**design-intelligence.json** -> `<session-folder>/analysis/design-intelligence.json` (UI mode only):
+- Produced when frontend keywords detected in seed_analysis
+- Fields: industry, style_direction, ux_patterns, color_strategy, typography, component_patterns
+- Consumed by architect (for design-tokens.json) and fe-developer
+
+### 4b: Inline Discuss (DISCUSS-001)
+
+After packaging, call discuss subagent:
+
+```
+Task({
+  subagent_type: "cli-discuss-agent",
+  run_in_background: false,
+  description: "Discuss DISCUSS-001",
+  prompt: `## Multi-Perspective Critique: DISCUSS-001
+
+### Input
+- Artifact: <session-folder>/spec/discovery-context.json
+- Round: DISCUSS-001
+- Perspectives: product, risk, coverage
+- Session: <session-folder>
+- Discovery Context: <session-folder>/spec/discovery-context.json
+
+<rest of discuss subagent prompt from subagents/discuss-subagent.md>`
+})
+```
+
+**Discuss result handling**:
+
+| Verdict | Severity | Action |
+|---------|----------|--------|
+| consensus_reached | - | Include action items in report, proceed to Phase 5 |
+| consensus_blocked | HIGH | Phase 5 SendMessage includes structured consensus_blocked format (see below). Do NOT self-revise. |
+| consensus_blocked | MEDIUM | Phase 5 SendMessage includes warning. Proceed normally. |
+| consensus_blocked | LOW | Treat as consensus_reached with notes. |
+
+**consensus_blocked SendMessage format**:
+```
+[analyst] RESEARCH-001 complete. Discuss DISCUSS-001: consensus_blocked (severity=<severity>)
+Divergences: <top-3-divergent-points>
+Action items: <prioritized-items>
+Recommendation: <revise|proceed-with-caution|escalate>
+Artifact: <session-folder>/spec/discovery-context.json
+Discussion: <session-folder>/discussions/DISCUSS-001-discussion.md
+```
+
+**Report**: complexity, codebase presence, problem statement, exploration dimensions, discuss verdict + severity, output paths.
+
+**Success**: Both JSON files created; discuss record written; design-intelligence.json created if UI mode.
+
+---
+
+## Error Handling
+
+| Scenario | Resolution |
+|----------|------------|
+| Gemini CLI failure | Fallback to direct Claude analysis |
+| Codebase detection failed | Continue as new project |
+| Topic too vague | Report with clarification questions |
+| Explore subagent fails | Continue without codebase context |
+| Discuss subagent fails | Proceed without discuss, log warning in report |

package/.claude/skills/team-lifecycle-v4/roles/architect/commands/assess.md

@@ -0,0 +1,193 @@
+# Command: assess
+
+## Purpose
+
+Multi-mode architecture assessment. Auto-detects consultation mode from task subject prefix, runs mode-specific analysis, and produces a verdict with concerns and recommendations.
+
+## Phase 2: Context Loading
+
+### Common Context (all modes)
+
+| Input | Source | Required |
+|-------|--------|----------|
+| Session folder | Task description `Session:` field | Yes |
+| Wisdom | `<session_folder>/wisdom/` (all files) | No |
+| Project tech | `.workflow/project-tech.json` | No |
+| Explorations | `<session_folder>/explorations/` | No |
+
+### Mode-Specific Context
+
+| Mode | Task Pattern | Additional Context |
+|------|-------------|-------------------|
+| spec-review | ARCH-SPEC-* | `spec/architecture/_index.md`, `spec/architecture/ADR-*.md` |
+| plan-review | ARCH-PLAN-* | `plan/plan.json`, `plan/.task/TASK-*.json` |
+| code-review | ARCH-CODE-* | `git diff --name-only`, changed file contents |
+| consult | ARCH-CONSULT-* | Question extracted from task description |
+| feasibility | ARCH-FEASIBILITY-* | Proposal from task description, codebase search results |
+
+## Phase 3: Mode-Specific Assessment
+
+### Mode: spec-review
+
+Review architecture documents for technical soundness across 4 dimensions.
+
+**Assessment dimensions**:
+
+| Dimension | Weight | Focus |
+|-----------|--------|-------|
+| Consistency | 25% | ADR decisions align with each other and with architecture index |
+| Scalability | 25% | Design supports growth, no single-point bottlenecks |
+| Security | 25% | Auth model, data protection, API security addressed |
+| Tech fitness | 25% | Technology choices match project-tech.json and problem domain |
+
+**Checks**:
+- Read architecture index and all ADR files
+- Cross-reference ADR decisions for contradictions
+- Verify tech choices align with project-tech.json
+- Score each dimension 0-100
+
+---
+
+### Mode: plan-review
+
+Review implementation plan for architectural soundness.
+
+**Checks**:
+
+| Check | What | Severity if Failed |
+|-------|------|-------------------|
+| Dependency cycles | Build task graph, detect cycles via DFS | High |
+| Task granularity | Flag tasks touching >8 files | Medium |
+| Convention compliance | Verify plan follows wisdom/conventions.md | Medium |
+| Architecture alignment | Verify plan doesn't contradict wisdom/decisions.md | High |
+
+**Dependency cycle detection flow**:
+1. Parse all TASK-*.json files -> extract id and depends_on
+2. Build directed graph
+3. DFS traversal -> flag any node visited twice in same stack
+4. Report cycle path if found
+
+---
+
+### Mode: code-review
+
+Assess architectural impact of code changes.
+
+**Checks**:
+
+| Check | Method | Severity if Found |
+|-------|--------|-------------------|
+| Layer violations | Detect upward imports (deeper layer importing shallower) | High |
+| New dependencies | Parse package.json diff for added deps | Medium |
+| Module boundary changes | Flag index.ts/index.js modifications | Medium |
+| Architectural impact | Score based on file count and boundary changes | Info |
+
+**Impact scoring**:
+
+| Condition | Impact Level |
+|-----------|-------------|
+| Changed files > 10 | High |
+| index.ts/index.js or package.json modified | Medium |
+| All other cases | Low |
+
+**Detection example** (find changed files):
+
+```bash
+Bash(command="git diff --name-only HEAD~1 2>/dev/null || git diff --name-only --cached")
+```
+
+---
+
+### Mode: consult
+
+Answer architecture decision questions. Route by question complexity.
+
+**Complexity detection**:
+
+| Condition | Classification |
+|-----------|---------------|
+| Question > 200 chars OR matches: architect, design, pattern, refactor, migrate, scalab | Complex |
+| All other questions | Simple |
+
+**Complex questions** -> delegate to CLI exploration:
+
+```bash
+Bash(command="ccw cli -p \"PURPOSE: Architecture consultation for: <question_summary>
+TASK: Search codebase for relevant patterns, analyze architectural implications, provide options with trade-offs
+MODE: analysis
+CONTEXT: @**/*
+EXPECTED: Options with trade-offs, file references, architectural implications
+CONSTRAINTS: Advisory only, provide options not decisions\" --tool gemini --mode analysis --rule analysis-review-architecture", timeout=300000)
+```
+
+**Simple questions** -> direct analysis using available context (wisdom, project-tech, codebase search).
+
+---
+
+### Mode: feasibility
+
+Evaluate technical feasibility of a proposal.
+
+**Assessment areas**:
+
+| Area | Method | Output |
+|------|--------|--------|
+| Tech stack compatibility | Compare proposal needs against project-tech.json | Compatible / Requires additions |
+| Codebase readiness | Search for integration points using Grep/Glob | Touch-point count |
+| Effort estimation | Based on touch-point count (see table below) | Low / Medium / High |
+| Risk assessment | Based on effort + tech compatibility | Risks + mitigations |
+
+**Effort estimation**:
+
+| Touch Points | Effort | Implication |
+|-------------|--------|-------------|
+| <= 5 | Low | Straightforward implementation |
+| 6 - 20 | Medium | Moderate refactoring needed |
+| > 20 | High | Significant refactoring, consider phasing |
+
+**Verdict for feasibility**:
+
+| Condition | Verdict |
+|-----------|---------|
+| Low/medium effort, compatible stack | FEASIBLE |
+| High touch-points OR new tech required | RISKY |
+| Fundamental incompatibility or unreasonable effort | INFEASIBLE |
+
+---
+
+### Verdict Routing (all modes except feasibility)
+
+| Verdict | Criteria |
+|---------|----------|
+| BLOCK | >= 2 high-severity concerns |
+| CONCERN | >= 1 high-severity OR >= 3 medium-severity concerns |
+| APPROVE | All other cases |
+
+## Phase 4: Validation
+
+### Output Format
+
+Write assessment to `<session_folder>/architecture/arch-<slug>.json`.
+
+**Report content sent to coordinator**:
+
+| Field | Description |
+|-------|-------------|
+| mode | Consultation mode used |
+| verdict | APPROVE / CONCERN / BLOCK (or FEASIBLE / RISKY / INFEASIBLE) |
+| concern_count | Number of concerns by severity |
+| recommendations | Actionable suggestions with trade-offs |
+| output_path | Path to full assessment file |
+
+**Wisdom contribution**: Append significant decisions to `<session_folder>/wisdom/decisions.md`.
+
+## Error Handling
+
+| Scenario | Resolution |
+|----------|------------|
+| Architecture docs not found | Assess from available context, note limitation in report |
+| Plan file missing | Report to coordinator via arch_concern |
+| Git diff fails (no commits) | Use staged changes or skip code-review mode |
+| CLI exploration timeout | Provide partial assessment, flag as incomplete |
+| Exploration results unparseable | Fall back to direct analysis without exploration |
+| Insufficient context | Request explorer assistance via coordinator |

package/.claude/skills/team-lifecycle-v4/roles/architect/role.md

@@ -0,0 +1,103 @@
+# Role: architect
+
+Architecture consultant. Advice on decisions, feasibility, design patterns.
+
+## Identity
+
+- **Name**: `architect` | **Prefix**: `ARCH-*` | **Tag**: `[architect]`
+- **Type**: Consulting (on-demand, advisory only)
+- **Responsibility**: Context loading → Mode detection → Analysis → Report
+
+## Boundaries
+
+### MUST
+- Only process ARCH-* tasks
+- Auto-detect mode from task subject prefix
+- Provide options with trade-offs (not final decisions)
+
+### MUST NOT
+- Modify source code
+- Make final decisions (advisory only)
+- Execute implementation or testing
+
+## Message Types
+
+| Type | Direction | Trigger |
+|------|-----------|---------|
+| arch_ready | → coordinator | Assessment complete |
+| arch_concern | → coordinator | Significant risk found |
+| error | → coordinator | Analysis failure |
+
+## Toolbox
+
+| Tool | Purpose |
+|------|---------|
+| commands/assess.md | Multi-mode assessment |
+| cli-explore-agent | Deep architecture exploration |
+| ccw cli --tool gemini --mode analysis | Architecture analysis |
+
+---
+
+## Consultation Modes
+
+| Task Pattern | Mode | Focus |
+|-------------|------|-------|
+| ARCH-SPEC-* | spec-review | Review architecture docs |
+| ARCH-PLAN-* | plan-review | Review plan soundness |
+| ARCH-CODE-* | code-review | Assess code change impact |
+| ARCH-CONSULT-* | consult | Answer architecture questions |
+| ARCH-FEASIBILITY-* | feasibility | Technical feasibility |
+
+---
+
+## Phase 2: Context Loading
+
+**Common**: session folder, wisdom, project-tech.json, explorations
+
+**Mode-specific**:
+
+| Mode | Additional Context |
+|------|-------------------|
+| spec-review | architecture/_index.md, ADR-*.md |
+| plan-review | plan/plan.json |
+| code-review | git diff, changed files |
+| consult | Question from task description |
+| feasibility | Requirements + codebase |
+
+---
+
+## Phase 3: Assessment
+
+Delegate to `commands/assess.md`. Output: mode, verdict (APPROVE/CONCERN/BLOCK), dimensions[], concerns[], recommendations[].
+
+For complex questions → Gemini CLI with architecture review rule.
+
+---
+
+## Phase 4: Report
+
+Output to `<session-folder>/architecture/arch-<slug>.json`. Contribute decisions to wisdom/decisions.md.
+
+**Frontend project outputs** (when frontend tech stack detected in shared-memory or discovery-context):
+- `<session-folder>/architecture/design-tokens.json` — color, spacing, typography, shadow tokens
+- `<session-folder>/architecture/component-specs/*.md` — per-component design spec
+
+**Report**: mode, verdict, concern count, recommendations, output path(s).
+
+---
+
+## Coordinator Integration
+
+| Timing | Task |
+|--------|------|
+| After DRAFT-003 | ARCH-SPEC-001: 架构文档评审 |
+| After PLAN-001 | ARCH-PLAN-001: 计划架构审查 |
+| On-demand | ARCH-CONSULT-001: 架构咨询 |
+
+## Error Handling
+
+| Scenario | Resolution |
+|----------|------------|
+| Docs not found | Assess from available context |
+| CLI timeout | Partial assessment |
+| Insufficient context | Request explorer via coordinator |

package/.claude/skills/team-lifecycle-v4/roles/coordinator/commands/dispatch.md

@@ -0,0 +1,198 @@
+# Command: dispatch
+
+## Purpose
+
+Create task chains based on execution mode. v4 optimized: spec pipeline reduced from 12 to 6 tasks by inlining discuss rounds into produce roles.
+
+## Phase 2: Context Loading
+
+| Input | Source | Required |
+|-------|--------|----------|
+| Mode | Phase 1 requirements (spec-only, impl-only, etc.) | Yes |
+| Session folder | `<session-folder>` from Phase 2 | Yes |
+| Scope | User requirements description | Yes |
+| Spec file | User-provided path (impl-only mode only) | Conditional |
+
+## Phase 3: Task Chain Creation
+
+### Mode-to-Pipeline Routing
+
+| Mode | Tasks | Pipeline | First Task |
+|------|-------|----------|------------|
+| spec-only | 6 | Spec pipeline | RESEARCH-001 |
+| impl-only | 4 | Impl pipeline | PLAN-001 |
+| fe-only | 3 | FE pipeline | PLAN-001 |
+| fullstack | 6 | Fullstack pipeline | PLAN-001 |
+| full-lifecycle | 10 | Spec + Impl | RESEARCH-001 |
+| full-lifecycle-fe | 12 | Spec + Fullstack | RESEARCH-001 |
+
+---
+
+### Spec Pipeline (6 tasks, was 12 in v3)
+
+Used by: spec-only, full-lifecycle, full-lifecycle-fe
+
+| # | Subject | Owner | BlockedBy | Description | Inline Discuss |
+|---|---------|-------|-----------|-------------|---------------|
+| 1 | RESEARCH-001 | analyst | (none) | Seed analysis and context gathering | DISCUSS-001 |
+| 2 | DRAFT-001 | writer | RESEARCH-001 | Generate Product Brief | DISCUSS-002 |
+| 3 | DRAFT-002 | writer | DRAFT-001 | Generate Requirements/PRD | DISCUSS-003 |
+| 4 | DRAFT-003 | writer | DRAFT-002 | Generate Architecture Document | DISCUSS-004 |
+| 5 | DRAFT-004 | writer | DRAFT-003 | Generate Epics & Stories | DISCUSS-005 |
+| 6 | QUALITY-001 | reviewer | DRAFT-004 | 5-dimension spec quality + sign-off | DISCUSS-006 |
+
+**Key change from v3**: No separate DISCUSS-* tasks. Each produce role calls the discuss subagent inline after completing its primary output.
+
+### Impl Pipeline (4 tasks)
+
+Used by: impl-only, full-lifecycle (PLAN-001 blockedBy QUALITY-001)
+
+| # | Subject | Owner | BlockedBy | Description |
+|---|---------|-------|-----------|-------------|
+| 1 | PLAN-001 | planner | (none) | Multi-angle exploration and planning |
+| 2 | IMPL-001 | executor | PLAN-001 | Code implementation |
+| 3 | TEST-001 | tester | IMPL-001 | Test-fix cycles |
+| 4 | REVIEW-001 | reviewer | IMPL-001 | 4-dimension code review |
+
+### FE Pipeline (3 tasks)
+
+Used by: fe-only
+
+| # | Subject | Owner | BlockedBy | Description |
+|---|---------|-------|-----------|-------------|
+| 1 | PLAN-001 | planner | (none) | Planning (frontend focus) |
+| 2 | DEV-FE-001 | fe-developer | PLAN-001 | Frontend implementation |
+| 3 | QA-FE-001 | fe-qa | DEV-FE-001 | 5-dimension frontend QA |
+
+GC loop (max 2 rounds): QA-FE verdict=NEEDS_FIX -> create DEV-FE-002 + QA-FE-002 dynamically.
+
+### Fullstack Pipeline (6 tasks)
+
+Used by: fullstack, full-lifecycle-fe (PLAN-001 blockedBy QUALITY-001)
+
+| # | Subject | Owner | BlockedBy | Description |
+|---|---------|-------|-----------|-------------|
+| 1 | PLAN-001 | planner | (none) | Fullstack planning |
+| 2 | IMPL-001 | executor | PLAN-001 | Backend implementation |
+| 3 | DEV-FE-001 | fe-developer | PLAN-001 | Frontend implementation |
+| 4 | TEST-001 | tester | IMPL-001 | Backend test-fix cycles |
+| 5 | QA-FE-001 | fe-qa | DEV-FE-001 | Frontend QA |
+| 6 | REVIEW-001 | reviewer | TEST-001, QA-FE-001 | Full code review |
+
+### Composite Modes
+
+| Mode | Construction | PLAN-001 BlockedBy |
+|------|-------------|-------------------|
+| full-lifecycle | Spec (6) + Impl (4) | QUALITY-001 |
+| full-lifecycle-fe | Spec (6) + Fullstack (6) | QUALITY-001 |
+
+---
+
+### Impl-Only Pre-check
+
+Before creating impl-only tasks, verify specification exists:
+
+```
+Spec exists?
+  +- YES -> read spec path -> proceed with task creation
+  +- NO -> error: "impl-only requires existing spec, use spec-only or full-lifecycle"
+```
+
+### Task Description Template
+
+Every task description includes session, scope, and inline discuss metadata:
+
+```
+TaskCreate({
+  subject: "<TASK-ID>",
+  owner: "<role>",
+  description: "<task description from pipeline table>\nSession: <session-folder>\nScope: <scope>\nInlineDiscuss: <DISCUSS-NNN or none>\nInnerLoop: <true|false>",
+  blockedBy: [<dependency-list>],
+  status: "pending"
+})
+```
+
+**InnerLoop Flag Rules**:
+
+| Role | InnerLoop |
+|------|-----------|
+| writer (DRAFT-*) | true |
+| planner (PLAN-*) | true |
+| executor (IMPL-*) | true |
+| analyst, tester, reviewer, architect, fe-developer, fe-qa | false |
+
+### Execution Method
+
+| Method | Behavior |
+|--------|----------|
+| sequential | One task active at a time; next activated after predecessor completes |
+| parallel | Tasks with all deps met run concurrently (e.g., TEST-001 + REVIEW-001) |
+
+## Phase 4: Validation
+
+| Check | Criteria |
+|-------|----------|
+| Task count | Matches mode total from routing table |
+| Dependencies | Every blockedBy references an existing task subject |
+| Owner assignment | Each task owner matches SKILL.md Role Registry prefix |
+| Session reference | Every task description contains `Session: <session-folder>` |
+| Inline discuss | Spec tasks have InlineDiscuss field matching round config |
+
+### Revision Task Template
+
+When handleRevise/handleFeedback creates revision tasks:
+
+```
+TaskCreate({
+  subject: "<ORIGINAL-ID>-R1",
+  owner: "<same-role-as-original>",
+  description: "<revision-type> revision of <ORIGINAL-ID>.\n
+    Session: <session-folder>\n
+    Original artifact: <artifact-path>\n
+    User feedback: <feedback-text or 'system-initiated'>\n
+    Revision scope: <targeted|full>\n
+    InlineDiscuss: <same-discuss-round-as-original>\n
+    InnerLoop: <true|false based on role>",
+  status: "pending",
+  blockedBy: [<predecessor-R1 if cascaded>]
+})
+```
+
+**Revision naming**: `<ORIGINAL-ID>-R1` (max 1 revision per task; second revision -> `-R2`; third -> escalate to user)
+
+**Cascade blockedBy chain example** (revise DRAFT-002):
+- DRAFT-002-R1 (no blockedBy)
+- DRAFT-003-R1 (blockedBy: DRAFT-002-R1)
+- DRAFT-004-R1 (blockedBy: DRAFT-003-R1)
+- QUALITY-001-R1 (blockedBy: DRAFT-004-R1)
+
+### Improvement Task Template
+
+When handleImprove creates improvement tasks:
+
+```
+TaskCreate({
+  subject: "IMPROVE-<dimension>-001",
+  owner: "writer",
+  description: "Quality improvement: <dimension>.\n
+    Session: <session-folder>\n
+    Current score: <X>%\n
+    Target: 80%\n
+    Readiness report: <session>/spec/readiness-report.md\n
+    Weak areas: <extracted-from-report>\n
+    Strategy: <from-dimension-strategy-table>\n
+    InnerLoop: true",
+  status: "pending"
+})
+```
+
+Improvement tasks are always followed by a QUALITY-001-R1 recheck (blockedBy: IMPROVE task).
+
+## Error Handling
+
+| Scenario | Resolution |
+|----------|------------|
+| Unknown mode | Reject with supported mode list |
+| Missing spec for impl-only | Error, suggest spec-only or full-lifecycle |
+| TaskCreate fails | Log error, report to user |
+| Duplicate task subject | Skip creation, log warning |