mirror of
https://github.com/catlog22/Claude-Code-Workflow.git
synced 2026-02-10 02:24:35 +08:00
rename: collaborative-plan → collaborative-plan-with-file
This commit is contained in:
catlog22
@@ -1,5 +1,5 @@
|
|||||||
---
|
---
|
||||||
name: workflow:collaborative-plan
|
name: workflow:collaborative-plan-with-file
|
||||||
description: Unified collaborative planning with dynamic requirement splitting, parallel sub-agent exploration/understanding/planning, and automatic merge. Each agent maintains process files for full traceability.
|
description: Unified collaborative planning with dynamic requirement splitting, parallel sub-agent exploration/understanding/planning, and automatic merge. Each agent maintains process files for full traceability.
|
||||||
argument-hint: "[-y|--yes] <task description> [--max-agents=5] [--depth=normal|deep] [--merge-rule=consensus|priority]"
|
argument-hint: "[-y|--yes] <task description> [--max-agents=5] [--depth=normal|deep] [--merge-rule=consensus|priority]"
|
||||||
allowed-tools: TodoWrite(*), Task(*), AskUserQuestion(*), Read(*), Bash(*), Write(*), Glob(*), Grep(*), mcp__ace-tool__search_context(*)
|
allowed-tools: TodoWrite(*), Task(*), AskUserQuestion(*), Read(*), Bash(*), Write(*), Glob(*), Grep(*), mcp__ace-tool__search_context(*)
|
||||||
@@ -15,12 +15,12 @@ When `--yes` or `-y`: Auto-approve splits, use default merge rule, skip confirma
|
|||||||
|
|
||||||
```bash
|
```bash
|
||||||
# Basic usage
|
# Basic usage
|
||||||
/workflow:collaborative-plan "Implement real-time notification system"
|
/workflow:collaborative-plan-with-file "Implement real-time notification system"
|
||||||
|
|
||||||
# With options
|
# With options
|
||||||
/workflow:collaborative-plan "Refactor authentication module" --max-agents=4
|
/workflow:collaborative-plan-with-file "Refactor authentication module" --max-agents=4
|
||||||
/workflow:collaborative-plan "Add payment gateway support" --depth=deep
|
/workflow:collaborative-plan-with-file "Add payment gateway support" --depth=deep
|
||||||
/workflow:collaborative-plan "Migrate to microservices" --merge-rule=priority
|
/workflow:collaborative-plan-with-file "Migrate to microservices" --merge-rule=priority
|
||||||
```
|
```
|
||||||
|
|
||||||
**Context Source**: ACE semantic search + Per-agent CLI exploration
|
**Context Source**: ACE semantic search + Per-agent CLI exploration
|
||||||
@@ -673,7 +673,7 @@ ${subRequirements.map(sub => `
|
|||||||
|
|
||||||
---
|
---
|
||||||
|
|
||||||
**Generated by**: /workflow:collaborative-plan
|
**Generated by**: /workflow:collaborative-plan-with-file
|
||||||
**Merge Rule**: ${mergeRule}
|
**Merge Rule**: ${mergeRule}
|
||||||
`
|
`
|
||||||
}
|
}
|
||||||
@@ -745,7 +745,7 @@ cat ${sessionFolder}/agents/{focus-area}/planning-context.md
|
|||||||
|
|
||||||
| Command | Use Case |
|
| Command | Use Case |
|
||||||
|---------|----------|
|
|---------|----------|
|
||||||
| **collaborative-plan** | Complex multi-aspect requirements needing parallel exploration |
|
| **collaborative-plan-with-file** | Complex multi-aspect requirements needing parallel exploration |
|
||||||
| lite-plan | Simple single-focus tasks |
|
| lite-plan | Simple single-focus tasks |
|
||||||
| multi-cli-plan | Iterative cross-verification with convergence |
|
| multi-cli-plan | Iterative cross-verification with convergence |
|
||||||
|
|
||||||
@@ -758,4 +758,4 @@ cat ${sessionFolder}/agents/{focus-area}/planning-context.md
|
|||||||
|
|
||||||
---
|
---
|
||||||
|
|
||||||
**Now execute collaborative-plan for**: $ARGUMENTS
|
**Now execute collaborative-plan-with-file for**: $ARGUMENTS
|
||||||
@@ -1,808 +0,0 @@
|
|||||||
---
|
|
||||||
name: quick-plan-with-file
|
|
||||||
description: Multi-agent rapid planning with minimal documentation, conflict resolution, and actionable synthesis. Designed as a lightweight planning supplement between brainstorm and full implementation planning
|
|
||||||
argument-hint: "[-y|--yes] [-c|--continue] [-f|--from <type>] \"planning topic or task description\""
|
|
||||||
allowed-tools: TodoWrite(*), Task(*), AskUserQuestion(*), Read(*), Grep(*), Glob(*), Bash(*), Edit(*), Write(*)
|
|
||||||
---
|
|
||||||
|
|
||||||
## Auto Mode
|
|
||||||
|
|
||||||
When `--yes` or `-y`: Auto-confirm planning decisions, use aggressive parallelization, minimal user interaction.
|
|
||||||
|
|
||||||
# Workflow Quick-Plan-With-File Command (/workflow:quick-plan-with-file)
|
|
||||||
|
|
||||||
## Overview
|
|
||||||
|
|
||||||
Multi-agent rapid planning workflow with **minimal documentation overhead**. Coordinates parallel agent analysis, synthesizes conflicting perspectives into actionable decisions, and generates a lightweight implementation-ready plan.
|
|
||||||
|
|
||||||
**Core workflow**: Parse Input → Parallel Analysis → Conflict Resolution → Plan Synthesis → Output
|
|
||||||
|
|
||||||
**Key features**:
|
|
||||||
- **Plan Format Agnostic**: Consumes brainstorm conclusions, analysis recommendations, or raw task descriptions
|
|
||||||
- **Minimal Docs**: Single `plan.md` (no lengthy brainstorm.md or discussion.md)
|
|
||||||
- **Parallel Multi-Agent**: 3-4 concurrent agent perspectives (architecture, implementation, validation, risk)
|
|
||||||
- **Conflict Resolution**: Automatic conflict detection and resolution via synthesis agent
|
|
||||||
- **Actionable Output**: Direct task breakdown ready for execution
|
|
||||||
- **Session Resumable**: Continue if interrupted, checkpoint at each phase
|
|
||||||
|
|
||||||
## Usage
|
|
||||||
|
|
||||||
```bash
|
|
||||||
/workflow:quick-plan-with-file [FLAGS] <PLANNING_TOPIC>
|
|
||||||
|
|
||||||
# Flags
|
|
||||||
-y, --yes Auto-confirm decisions, use defaults
|
|
||||||
-c, --continue Continue existing session (auto-detected)
|
|
||||||
-f, --from <type> Input source type: brainstorm|analysis|task|raw
|
|
||||||
|
|
||||||
# Arguments
|
|
||||||
<planning-topic> Planning topic, task, or reference to planning artifact
|
|
||||||
|
|
||||||
# Examples
|
|
||||||
/workflow:quick-plan-with-file "实现分布式缓存层，支持Redis和内存后端"
|
|
||||||
/workflow:quick-plan-with-file --continue "缓存层规划" # Continue
|
|
||||||
/workflow:quick-plan-with-file -y -f analysis "从分析结论生成实施规划" # Auto mode
|
|
||||||
/workflow:quick-plan-with-file --from brainstorm BS-rate-limiting-2025-01-28 # From artifact
|
|
||||||
```
|
|
||||||
|
|
||||||
## Execution Process
|
|
||||||
|
|
||||||
```
|
|
||||||
Input Validation & Loading:
|
|
||||||
├─ Parse input (topic | artifact reference)
|
|
||||||
├─ Load artifact if referenced (synthesis.json | conclusions.json | etc.)
|
|
||||||
├─ Extract key constraints and requirements
|
|
||||||
└─ Initialize session folder and plan.md
|
|
||||||
|
|
||||||
Session Initialization:
|
|
||||||
├─ Create .workflow/.planning/{sessionId}/
|
|
||||||
├─ Initialize plan.md with input summary
|
|
||||||
├─ Parse existing output (if --from artifact)
|
|
||||||
└─ Define planning dimensions & focus areas
|
|
||||||
|
|
||||||
Phase 1: Parallel Multi-Agent Analysis (concurrent)
|
|
||||||
├─ Agent 1 (Architecture): High-level design & decomposition
|
|
||||||
├─ Agent 2 (Implementation): Technical approach & feasibility
|
|
||||||
├─ Agent 3 (Validation): Risk analysis & edge cases
|
|
||||||
├─ Agent 4 (Decision): Recommendations & tradeoffs
|
|
||||||
└─ Aggregate findings into perspectives.json
|
|
||||||
|
|
||||||
Phase 2: Conflict Detection & Resolution
|
|
||||||
├─ Analyze agent perspectives for contradictions
|
|
||||||
├─ Identify critical decision points
|
|
||||||
├─ Generate synthesis via arbitration agent
|
|
||||||
├─ Document conflicts and resolutions
|
|
||||||
└─ Update plan.md with decisive recommendations
|
|
||||||
|
|
||||||
Phase 3: Plan Synthesis
|
|
||||||
├─ Consolidate all insights
|
|
||||||
├─ Generate actionable task breakdown
|
|
||||||
├─ Create execution strategy
|
|
||||||
├─ Document assumptions & risks
|
|
||||||
└─ Generate synthesis.md with ready-to-execute tasks
|
|
||||||
|
|
||||||
Output:
|
|
||||||
├─ .workflow/.planning/{sessionId}/plan.md (minimal, actionable)
|
|
||||||
├─ .workflow/.planning/{sessionId}/perspectives.json (agent findings)
|
|
||||||
├─ .workflow/.planning/{sessionId}/conflicts.json (decision points)
|
|
||||||
├─ .workflow/.planning/{sessionId}/synthesis.md (task breakdown)
|
|
||||||
└─ Optional: Feed to /workflow:unified-execute-with-file
|
|
||||||
```
|
|
||||||
|
|
||||||
## Implementation
|
|
||||||
|
|
||||||
### Session Setup & Input Loading
|
|
||||||
|
|
||||||
```javascript
|
|
||||||
const getUtc8ISOString = () => new Date(Date.now() + 8 * 60 * 60 * 1000).toISOString()
|
|
||||||
|
|
||||||
// Parse arguments
|
|
||||||
const planningTopic = "$PLANNING_TOPIC"
|
|
||||||
const inputType = $ARGUMENTS.match(/--from\s+(\w+)/)?.[1] || 'raw'
|
|
||||||
const isAutoMode = $ARGUMENTS.includes('--yes') || $ARGUMENTS.includes('-y')
|
|
||||||
const isContinue = $ARGUMENTS.includes('--continue') || $ARGUMENTS.includes('-c')
|
|
||||||
|
|
||||||
// Auto-detect artifact if referenced
|
|
||||||
let artifact = null
|
|
||||||
let artifactContent = null
|
|
||||||
|
|
||||||
if (inputType === 'brainstorm' || planningTopic.startsWith('BS-')) {
|
|
||||||
const sessionId = planningTopic
|
|
||||||
const synthesisPath = `.workflow/.brainstorm/${sessionId}/synthesis.json`
|
|
||||||
if (fs.existsSync(synthesisPath)) {
|
|
||||||
artifact = { type: 'brainstorm', path: synthesisPath }
|
|
||||||
artifactContent = JSON.parse(Read(synthesisPath))
|
|
||||||
}
|
|
||||||
} else if (inputType === 'analysis' || planningTopic.startsWith('ANL-')) {
|
|
||||||
const sessionId = planningTopic
|
|
||||||
const conclusionsPath = `.workflow/.analysis/${sessionId}/conclusions.json`
|
|
||||||
if (fs.existsSync(conclusionsPath)) {
|
|
||||||
artifact = { type: 'analysis', path: conclusionsPath }
|
|
||||||
artifactContent = JSON.parse(Read(conclusionsPath))
|
|
||||||
}
|
|
||||||
}
|
|
||||||
|
|
||||||
// Generate session ID
|
|
||||||
const planSlug = planningTopic.toLowerCase()
|
|
||||||
.replace(/[^a-z0-9\u4e00-\u9fa5]+/g, '-')
|
|
||||||
.substring(0, 30)
|
|
||||||
const dateStr = getUtc8ISOString().substring(0, 10)
|
|
||||||
const sessionId = `PLAN-${planSlug}-${dateStr}`
|
|
||||||
const sessionFolder = `.workflow/.planning/${sessionId}`
|
|
||||||
|
|
||||||
// Session mode detection
|
|
||||||
const sessionExists = fs.existsSync(sessionFolder)
|
|
||||||
const hasPlan = sessionExists && fs.existsSync(`${sessionFolder}/plan.md`)
|
|
||||||
const mode = (hasPlan || isContinue) ? 'continue' : 'new'
|
|
||||||
|
|
||||||
if (!sessionExists) {
|
|
||||||
bash(`mkdir -p ${sessionFolder}`)
|
|
||||||
}
|
|
||||||
```
|
|
||||||
|
|
||||||
---
|
|
||||||
|
|
||||||
### Phase 1: Initialize plan.md (Minimal)
|
|
||||||
|
|
||||||
```markdown
|
|
||||||
# Quick Planning Session
|
|
||||||
|
|
||||||
**Session ID**: ${sessionId}
|
|
||||||
**Topic**: ${planningTopic}
|
|
||||||
**Started**: ${getUtc8ISOString()}
|
|
||||||
**Mode**: ${mode}
|
|
||||||
|
|
||||||
---
|
|
||||||
|
|
||||||
## Input Context
|
|
||||||
|
|
||||||
${artifact ? `
|
|
||||||
**Source**: ${artifact.type} artifact
|
|
||||||
**Path**: ${artifact.path}
|
|
||||||
|
|
||||||
**Artifact Summary**:
|
|
||||||
${artifact.type === 'brainstorm' ? `
|
|
||||||
- Topic: ${artifactContent.topic}
|
|
||||||
- Top Ideas: ${artifactContent.top_ideas?.length || 0}
|
|
||||||
- Key Insights: ${artifactContent.key_insights?.slice(0, 2).join(', ') || 'N/A'}
|
|
||||||
` : artifact.type === 'analysis' ? `
|
|
||||||
- Topic: ${artifactContent.topic}
|
|
||||||
- Key Conclusions: ${artifactContent.key_conclusions?.length || 0}
|
|
||||||
- Recommendations: ${artifactContent.recommendations?.length || 0}
|
|
||||||
` : ''}
|
|
||||||
` : `
|
|
||||||
**User Input**: ${planningTopic}
|
|
||||||
`}
|
|
||||||
|
|
||||||
---
|
|
||||||
|
|
||||||
## Planning Dimensions
|
|
||||||
|
|
||||||
*To be populated after agent analysis*
|
|
||||||
|
|
||||||
---
|
|
||||||
|
|
||||||
## Key Decisions
|
|
||||||
|
|
||||||
*Conflict resolution and recommendations - to be populated*
|
|
||||||
|
|
||||||
---
|
|
||||||
|
|
||||||
## Implementation Plan
|
|
||||||
|
|
||||||
*Task breakdown - to be populated after synthesis*
|
|
||||||
|
|
||||||
---
|
|
||||||
|
|
||||||
## Progress
|
|
||||||
|
|
||||||
- [ ] Multi-agent analysis
|
|
||||||
- [ ] Conflict detection
|
|
||||||
- [ ] Plan synthesis
|
|
||||||
- [ ] Ready for execution
|
|
||||||
```
|
|
||||||
|
|
||||||
---
|
|
||||||
|
|
||||||
### Phase 2: Parallel Multi-Agent Analysis
|
|
||||||
|
|
||||||
```javascript
|
|
||||||
const analysisPrompt = artifact
|
|
||||||
? `Convert ${artifact.type} artifact to planning requirements and execute parallel analysis`
|
|
||||||
: `Create planning breakdown for: ${planningTopic}`
|
|
||||||
|
|
||||||
// Prepare context for agents
|
|
||||||
const agentContext = {
|
|
||||||
topic: planningTopic,
|
|
||||||
artifact: artifact ? {
|
|
||||||
type: artifact.type,
|
|
||||||
summary: extractArtifactSummary(artifactContent)
|
|
||||||
} : null,
|
|
||||||
planning_focus: determineFocusAreas(planningTopic),
|
|
||||||
constraints: extractConstraints(planningTopic, artifactContent)
|
|
||||||
}
|
|
||||||
|
|
||||||
// Agent 1: Architecture & Design
|
|
||||||
const archPromise = Bash({
|
|
||||||
command: `ccw cli -p "
|
|
||||||
PURPOSE: Architecture & high-level design planning for '${planningTopic}'
|
|
||||||
Success: Clear component decomposition, interface design, and data flow
|
|
||||||
|
|
||||||
TASK:
|
|
||||||
• Decompose problem into major components/modules
|
|
||||||
• Identify architectural patterns and integration points
|
|
||||||
• Design interfaces and data models
|
|
||||||
• Assess scalability and maintainability implications
|
|
||||||
• Propose architectural approach with rationale
|
|
||||||
|
|
||||||
MODE: analysis
|
|
||||||
|
|
||||||
CONTEXT: @**/*
|
|
||||||
${artifact ? `| Source artifact: ${artifact.type}` : ''}
|
|
||||||
|
|
||||||
EXPECTED:
|
|
||||||
- Component decomposition (box diagram in text)
|
|
||||||
- Module interfaces and responsibilities
|
|
||||||
- Data flow between components
|
|
||||||
- Architectural patterns applied
|
|
||||||
- Scalability assessment (1-5 rating)
|
|
||||||
- Risks from architectural perspective
|
|
||||||
|
|
||||||
CONSTRAINTS: Focus on long-term maintainability
|
|
||||||
" --tool gemini --mode analysis`,
|
|
||||||
run_in_background: true
|
|
||||||
})
|
|
||||||
|
|
||||||
// Agent 2: Implementation & Feasibility
|
|
||||||
const implPromise = Bash({
|
|
||||||
command: `ccw cli -p "
|
|
||||||
PURPOSE: Implementation approach & technical feasibility for '${planningTopic}'
|
|
||||||
Success: Concrete implementation strategy with realistic resource estimates
|
|
||||||
|
|
||||||
TASK:
|
|
||||||
• Evaluate technical feasibility of approach
|
|
||||||
• Identify required technologies and dependencies
|
|
||||||
• Estimate effort: high/medium/low + rationale
|
|
||||||
• Suggest implementation phases and milestones
|
|
||||||
• Highlight technical blockers or challenges
|
|
||||||
|
|
||||||
MODE: analysis
|
|
||||||
|
|
||||||
CONTEXT: @**/*
|
|
||||||
${artifact ? `| Source artifact: ${artifact.type}` : ''}
|
|
||||||
|
|
||||||
EXPECTED:
|
|
||||||
- Technology stack recommendation
|
|
||||||
- Implementation complexity: high|medium|low with justification
|
|
||||||
- Estimated effort breakdown (analysis/design/coding/testing/deployment)
|
|
||||||
- Key technical decisions with tradeoffs
|
|
||||||
- Potential blockers and mitigations
|
|
||||||
- Suggested implementation phases
|
|
||||||
- Reusable components or libraries
|
|
||||||
|
|
||||||
CONSTRAINTS: Realistic with current tech stack
|
|
||||||
" --tool codex --mode analysis`,
|
|
||||||
run_in_background: true
|
|
||||||
})
|
|
||||||
|
|
||||||
// Agent 3: Risk & Validation
|
|
||||||
const riskPromise = Bash({
|
|
||||||
command: `ccw cli -p "
|
|
||||||
PURPOSE: Risk analysis and validation strategy for '${planningTopic}'
|
|
||||||
Success: Comprehensive risk matrix with testing strategy
|
|
||||||
|
|
||||||
TASK:
|
|
||||||
• Identify technical risks and failure scenarios
|
|
||||||
• Assess business/timeline risks
|
|
||||||
• Define validation/testing strategy
|
|
||||||
• Suggest monitoring and observability requirements
|
|
||||||
• Rate overall risk level (low/medium/high)
|
|
||||||
|
|
||||||
MODE: analysis
|
|
||||||
|
|
||||||
CONTEXT: @**/*
|
|
||||||
${artifact ? `| Source artifact: ${artifact.type}` : ''}
|
|
||||||
|
|
||||||
EXPECTED:
|
|
||||||
- Risk matrix (likelihood × impact, 1-5 each)
|
|
||||||
- Top 3 technical risks with mitigations
|
|
||||||
- Top 3 timeline/resource risks with mitigations
|
|
||||||
- Testing strategy (unit/integration/e2e/performance)
|
|
||||||
- Deployment strategy and rollback plan
|
|
||||||
- Monitoring/observability requirements
|
|
||||||
- Overall risk rating with confidence (low/medium/high)
|
|
||||||
|
|
||||||
CONSTRAINTS: Be realistic, not pessimistic
|
|
||||||
" --tool claude --mode analysis`,
|
|
||||||
run_in_background: true
|
|
||||||
})
|
|
||||||
|
|
||||||
// Agent 4: Decisions & Recommendations
|
|
||||||
const decisionPromise = Bash({
|
|
||||||
command: `ccw cli -p "
|
|
||||||
PURPOSE: Strategic decisions and execution recommendations for '${planningTopic}'
|
|
||||||
Success: Clear recommended approach with tradeoff analysis
|
|
||||||
|
|
||||||
TASK:
|
|
||||||
• Synthesize all considerations into recommendations
|
|
||||||
• Clearly identify critical decision points
|
|
||||||
• Outline key tradeoffs (speed vs quality, scope vs timeline, etc.)
|
|
||||||
• Propose go/no-go decision criteria
|
|
||||||
• Suggest execution strategy and sequencing
|
|
||||||
|
|
||||||
MODE: analysis
|
|
||||||
|
|
||||||
CONTEXT: @**/*
|
|
||||||
${artifact ? `| Source artifact: ${artifact.type}` : ''}
|
|
||||||
|
|
||||||
EXPECTED:
|
|
||||||
- Primary recommendation with strong rationale
|
|
||||||
- Alternative approaches with pros/cons
|
|
||||||
- 2-3 critical decision points with recommended choices
|
|
||||||
- Key tradeoffs and what we're optimizing for
|
|
||||||
- Success metrics and go/no-go criteria
|
|
||||||
- Suggested execution sequencing
|
|
||||||
- Resource requirements and dependencies
|
|
||||||
|
|
||||||
CONSTRAINTS: Focus on actionable decisions, not analysis
|
|
||||||
" --tool gemini --mode analysis`,
|
|
||||||
run_in_background: true
|
|
||||||
})
|
|
||||||
|
|
||||||
// Wait for all parallel analyses
|
|
||||||
const [archResult, implResult, riskResult, decisionResult] = await Promise.all([
|
|
||||||
archPromise, implPromise, riskPromise, decisionPromise
|
|
||||||
])
|
|
||||||
```
|
|
||||||
|
|
||||||
---
|
|
||||||
|
|
||||||
### Phase 3: Aggregate Perspectives
|
|
||||||
|
|
||||||
```javascript
|
|
||||||
// Parse and structure agent findings
|
|
||||||
const perspectives = {
|
|
||||||
session_id: sessionId,
|
|
||||||
timestamp: getUtc8ISOString(),
|
|
||||||
topic: planningTopic,
|
|
||||||
source_artifact: artifact?.type || 'raw',
|
|
||||||
|
|
||||||
architecture: {
|
|
||||||
source: 'gemini (design)',
|
|
||||||
components: extractComponents(archResult),
|
|
||||||
interfaces: extractInterfaces(archResult),
|
|
||||||
patterns: extractPatterns(archResult),
|
|
||||||
scalability_rating: extractRating(archResult, 'scalability'),
|
|
||||||
risks_from_design: extractRisks(archResult)
|
|
||||||
},
|
|
||||||
|
|
||||||
implementation: {
|
|
||||||
source: 'codex (pragmatic)',
|
|
||||||
technology_stack: extractStack(implResult),
|
|
||||||
complexity: extractComplexity(implResult),
|
|
||||||
effort_breakdown: extractEffort(implResult),
|
|
||||||
blockers: extractBlockers(implResult),
|
|
||||||
phases: extractPhases(implResult)
|
|
||||||
},
|
|
||||||
|
|
||||||
validation: {
|
|
||||||
source: 'claude (systematic)',
|
|
||||||
risk_matrix: extractRiskMatrix(riskResult),
|
|
||||||
top_risks: extractTopRisks(riskResult),
|
|
||||||
testing_strategy: extractTestingStrategy(riskResult),
|
|
||||||
deployment_strategy: extractDeploymentStrategy(riskResult),
|
|
||||||
monitoring_requirements: extractMonitoring(riskResult),
|
|
||||||
overall_risk_rating: extractRiskRating(riskResult)
|
|
||||||
},
|
|
||||||
|
|
||||||
recommendation: {
|
|
||||||
source: 'gemini (synthesis)',
|
|
||||||
primary_approach: extractPrimaryApproach(decisionResult),
|
|
||||||
alternatives: extractAlternatives(decisionResult),
|
|
||||||
critical_decisions: extractDecisions(decisionResult),
|
|
||||||
tradeoffs: extractTradeoffs(decisionResult),
|
|
||||||
success_criteria: extractCriteria(decisionResult),
|
|
||||||
execution_sequence: extractSequence(decisionResult)
|
|
||||||
},
|
|
||||||
|
|
||||||
analysis_timestamp: getUtc8ISOString()
|
|
||||||
}
|
|
||||||
|
|
||||||
Write(`${sessionFolder}/perspectives.json`, JSON.stringify(perspectives, null, 2))
|
|
||||||
```
|
|
||||||
|
|
||||||
---
|
|
||||||
|
|
||||||
### Phase 4: Conflict Detection & Resolution
|
|
||||||
|
|
||||||
```javascript
|
|
||||||
// Analyze for conflicts and contradictions
|
|
||||||
const conflicts = detectConflicts({
|
|
||||||
arch_vs_impl: compareArchitectureAndImplementation(perspectives),
|
|
||||||
design_vs_risk: compareDesignAndRisk(perspectives),
|
|
||||||
effort_vs_scope: compareEffortAndScope(perspectives),
|
|
||||||
timeline_implications: extractTimingConflicts(perspectives)
|
|
||||||
})
|
|
||||||
|
|
||||||
// If conflicts exist, invoke arbitration agent
|
|
||||||
if (conflicts.critical.length > 0) {
|
|
||||||
const arbitrationResult = await Bash({
|
|
||||||
command: `ccw cli -p "
|
|
||||||
PURPOSE: Resolve planning conflicts and generate unified recommendation
|
|
||||||
Input: ${JSON.stringify(conflicts, null, 2)}
|
|
||||||
|
|
||||||
TASK:
|
|
||||||
• Review all conflicts presented
|
|
||||||
• Recommend resolution for each critical conflict
|
|
||||||
• Explain tradeoff choices
|
|
||||||
• Identify what we're optimizing for (speed/quality/risk/resource)
|
|
||||||
• Generate unified execution strategy
|
|
||||||
|
|
||||||
MODE: analysis
|
|
||||||
|
|
||||||
EXPECTED:
|
|
||||||
- For each conflict: recommended resolution + rationale
|
|
||||||
- Unified optimization criteria (what matters most?)
|
|
||||||
- Final recommendation with confidence level
|
|
||||||
- Any unresolved tensions that need user input
|
|
||||||
|
|
||||||
CONSTRAINTS: Be decisive, not fence-sitting
|
|
||||||
" --tool gemini --mode analysis`,
|
|
||||||
run_in_background: false
|
|
||||||
})
|
|
||||||
|
|
||||||
const conflictResolution = {
|
|
||||||
detected_conflicts: conflicts,
|
|
||||||
arbitration_result: arbitrationResult,
|
|
||||||
timestamp: getUtc8ISOString()
|
|
||||||
}
|
|
||||||
|
|
||||||
Write(`${sessionFolder}/conflicts.json`, JSON.stringify(conflictResolution, null, 2))
|
|
||||||
}
|
|
||||||
```
|
|
||||||
|
|
||||||
---
|
|
||||||
|
|
||||||
### Phase 5: Plan Synthesis & Task Breakdown
|
|
||||||
|
|
||||||
```javascript
|
|
||||||
const synthesisPrompt = `
|
|
||||||
Given the planning context:
|
|
||||||
- Topic: ${planningTopic}
|
|
||||||
- Architecture: ${perspectives.architecture.components.map(c => c.name).join(', ')}
|
|
||||||
- Implementation Complexity: ${perspectives.implementation.complexity}
|
|
||||||
- Timeline Risk: ${perspectives.validation.overall_risk_rating}
|
|
||||||
- Primary Recommendation: ${perspectives.recommendation.primary_approach.summary}
|
|
||||||
|
|
||||||
Generate a minimal but complete implementation plan with:
|
|
||||||
1. Task breakdown (5-8 major tasks)
|
|
||||||
2. Dependencies between tasks
|
|
||||||
3. For each task: what needs to be done, why, and key considerations
|
|
||||||
4. Success criteria for the entire effort
|
|
||||||
5. Known risks and mitigation strategies
|
|
||||||
|
|
||||||
Output as structured task list ready for execution.
|
|
||||||
`
|
|
||||||
|
|
||||||
const synthesisResult = await Bash({
|
|
||||||
command: `ccw cli -p "${synthesisPrompt}" --tool gemini --mode analysis`,
|
|
||||||
run_in_background: false
|
|
||||||
})
|
|
||||||
|
|
||||||
// Parse synthesis and generate task breakdown
|
|
||||||
const tasks = parseTaskBreakdown(synthesisResult)
|
|
||||||
|
|
||||||
const synthesis = {
|
|
||||||
session_id: sessionId,
|
|
||||||
planning_topic: planningTopic,
|
|
||||||
completed: getUtc8ISOString(),
|
|
||||||
|
|
||||||
// Summary
|
|
||||||
executive_summary: perspectives.recommendation.primary_approach.summary,
|
|
||||||
optimization_focus: extractOptimizationFocus(perspectives),
|
|
||||||
|
|
||||||
// Architecture
|
|
||||||
architecture_approach: perspectives.architecture.patterns[0] || 'TBD',
|
|
||||||
key_components: perspectives.architecture.components.slice(0, 5),
|
|
||||||
|
|
||||||
// Implementation
|
|
||||||
technology_stack: perspectives.implementation.technology_stack,
|
|
||||||
complexity_level: perspectives.implementation.complexity,
|
|
||||||
estimated_effort: perspectives.implementation.effort_breakdown,
|
|
||||||
|
|
||||||
// Risks & Validation
|
|
||||||
top_risks: perspectives.validation.top_risks.slice(0, 3),
|
|
||||||
testing_approach: perspectives.validation.testing_strategy,
|
|
||||||
|
|
||||||
// Execution
|
|
||||||
phases: perspectives.implementation.phases,
|
|
||||||
critical_path_tasks: extractCriticalPath(tasks),
|
|
||||||
total_tasks: tasks.length,
|
|
||||||
|
|
||||||
// Task breakdown (ready for unified-execute-with-file)
|
|
||||||
tasks: tasks.map(task => ({
|
|
||||||
id: task.id,
|
|
||||||
title: task.title,
|
|
||||||
description: task.description,
|
|
||||||
type: task.type,
|
|
||||||
dependencies: task.dependencies,
|
|
||||||
effort_estimate: task.effort,
|
|
||||||
success_criteria: task.criteria
|
|
||||||
}))
|
|
||||||
}
|
|
||||||
|
|
||||||
Write(`${sessionFolder}/synthesis.md`, formatSynthesisMarkdown(synthesis))
|
|
||||||
Write(`${sessionFolder}/synthesis.json`, JSON.stringify(synthesis, null, 2))
|
|
||||||
```
|
|
||||||
|
|
||||||
---
|
|
||||||
|
|
||||||
### Phase 6: Update plan.md with Results
|
|
||||||
|
|
||||||
```markdown
|
|
||||||
# Quick Planning Session
|
|
||||||
|
|
||||||
**Session ID**: ${sessionId}
|
|
||||||
**Topic**: ${planningTopic}
|
|
||||||
**Started**: ${startTime}
|
|
||||||
**Completed**: ${completionTime}
|
|
||||||
|
|
||||||
---
|
|
||||||
|
|
||||||
## Executive Summary
|
|
||||||
|
|
||||||
${synthesis.executive_summary}
|
|
||||||
|
|
||||||
**Optimization Focus**: ${synthesis.optimization_focus}
|
|
||||||
**Complexity**: ${synthesis.complexity_level}
|
|
||||||
**Estimated Effort**: ${formatEffort(synthesis.estimated_effort)}
|
|
||||||
|
|
||||||
---
|
|
||||||
|
|
||||||
## Architecture
|
|
||||||
|
|
||||||
**Primary Pattern**: ${synthesis.architecture_approach}
|
|
||||||
|
|
||||||
**Key Components**:
|
|
||||||
${synthesis.key_components.map((c, i) => `${i+1}. ${c.name}: ${c.responsibility}`).join('\n')}
|
|
||||||
|
|
||||||
---
|
|
||||||
|
|
||||||
## Implementation Strategy
|
|
||||||
|
|
||||||
**Technology Stack**:
|
|
||||||
${synthesis.technology_stack.map(t => `- ${t}`).join('\n')}
|
|
||||||
|
|
||||||
**Phases**:
|
|
||||||
${synthesis.phases.map((p, i) => `${i+1}. ${p.name} (${p.effort})`).join('\n')}
|
|
||||||
|
|
||||||
---
|
|
||||||
|
|
||||||
## Risk Assessment
|
|
||||||
|
|
||||||
**Overall Risk Level**: ${synthesis.top_risks[0].risk_level}
|
|
||||||
|
|
||||||
**Top 3 Risks**:
|
|
||||||
${synthesis.top_risks.map((r, i) => `
|
|
||||||
${i+1}. **${r.title}** (Impact: ${r.impact})
|
|
||||||
- Mitigation: ${r.mitigation}
|
|
||||||
`).join('\n')}
|
|
||||||
|
|
||||||
**Testing Approach**: ${synthesis.testing_approach}
|
|
||||||
|
|
||||||
---
|
|
||||||
|
|
||||||
## Execution Plan
|
|
||||||
|
|
||||||
**Total Tasks**: ${synthesis.total_tasks}
|
|
||||||
**Critical Path**: ${synthesis.critical_path_tasks.map(t => t.id).join(' → ')}
|
|
||||||
|
|
||||||
### Task Breakdown
|
|
||||||
|
|
||||||
${synthesis.tasks.map((task, i) => `
|
|
||||||
${i+1}. **${task.id}: ${task.title}** (Effort: ${task.effort_estimate})
|
|
||||||
- ${task.description}
|
|
||||||
- Depends on: ${task.dependencies.join(', ') || 'none'}
|
|
||||||
- Success: ${task.success_criteria}
|
|
||||||
`).join('\n')}
|
|
||||||
|
|
||||||
---
|
|
||||||
|
|
||||||
## Next Steps
|
|
||||||
|
|
||||||
**Recommended**: Execute with \`/workflow:unified-execute-with-file\` using:
|
|
||||||
\`\`\`
|
|
||||||
/workflow:unified-execute-with-file -p ${sessionFolder}/synthesis.json
|
|
||||||
\`\`\`
|
|
||||||
|
|
||||||
---
|
|
||||||
|
|
||||||
## Artifacts
|
|
||||||
|
|
||||||
- **Perspectives**: ${sessionFolder}/perspectives.json (all agent findings)
|
|
||||||
- **Conflicts**: ${sessionFolder}/conflicts.json (decision points and resolutions)
|
|
||||||
- **Synthesis**: ${sessionFolder}/synthesis.json (task breakdown for execution)
|
|
||||||
```
|
|
||||||
|
|
||||||
---
|
|
||||||
|
|
||||||
## Session Folder Structure
|
|
||||||
|
|
||||||
```
|
|
||||||
.workflow/.planning/{sessionId}/
|
|
||||||
├── plan.md # Minimal, actionable planning doc
|
|
||||||
├── perspectives.json # Multi-agent findings (architecture, impl, risk, decision)
|
|
||||||
├── conflicts.json # Detected conflicts and resolutions (if any)
|
|
||||||
├── synthesis.json # Task breakdown ready for execution
|
|
||||||
└── synthesis.md # Human-readable execution plan
|
|
||||||
```
|
|
||||||
|
|
||||||
---
|
|
||||||
|
|
||||||
## Multi-Agent Roles
|
|
||||||
|
|
||||||
| Agent | Focus | Input | Output |
|
|
||||||
|-------|-------|-------|--------|
|
|
||||||
| **Gemini (Design)** | Architecture & design patterns | Topic + constraints | Components, interfaces, patterns, scalability |
|
|
||||||
| **Codex (Pragmatic)** | Implementation reality | Topic + architecture | Tech stack, effort, phases, blockers |
|
|
||||||
| **Claude (Validation)** | Risk & testing | Architecture + impl | Risk matrix, test strategy, monitoring |
|
|
||||||
| **Gemini (Decision)** | Synthesis & strategy | All findings | Recommendations, tradeoffs, execution plan |
|
|
||||||
|
|
||||||
---
|
|
||||||
|
|
||||||
## Conflict Resolution Strategy
|
|
||||||
|
|
||||||
**Auto-Resolution for conflicts**:
|
|
||||||
1. **Architecture vs Implementation**: Recommend design-for-feasibility approach
|
|
||||||
2. **Scope vs Timeline**: Prioritize critical path, defer nice-to-haves
|
|
||||||
3. **Quality vs Speed**: Suggest iterative approach (MVP + iterations)
|
|
||||||
4. **Resource vs Effort**: Identify parallelizable tasks
|
|
||||||
|
|
||||||
**Require User Input for**:
|
|
||||||
- Strategic choices (which feature to prioritize?)
|
|
||||||
- Tool/technology decisions with strong team preferences
|
|
||||||
- Budget/resource constraints not stated in planning topic
|
|
||||||
|
|
||||||
---
|
|
||||||
|
|
||||||
## Continue & Resume
|
|
||||||
|
|
||||||
```bash
|
|
||||||
/workflow:quick-plan-with-file --continue "planning-topic"
|
|
||||||
```
|
|
||||||
|
|
||||||
When continuing:
|
|
||||||
1. Load existing plan.md and perspectives.json
|
|
||||||
2. Identify what's incomplete
|
|
||||||
3. Re-run affected agents (if planning has changed)
|
|
||||||
4. Update plan.md with new findings
|
|
||||||
5. Generate updated synthesis.json
|
|
||||||
|
|
||||||
---
|
|
||||||
|
|
||||||
## Integration Flow
|
|
||||||
|
|
||||||
```
|
|
||||||
Input Source:
|
|
||||||
├─ Raw task description
|
|
||||||
├─ Brainstorm synthesis.json
|
|
||||||
└─ Analysis conclusions.json
|
|
||||||
↓
|
|
||||||
/workflow:quick-plan-with-file
|
|
||||||
↓
|
|
||||||
plan.md + synthesis.json
|
|
||||||
↓
|
|
||||||
/workflow:unified-execute-with-file
|
|
||||||
↓
|
|
||||||
Implementation
|
|
||||||
```
|
|
||||||
|
|
||||||
---
|
|
||||||
|
|
||||||
## Usage Patterns
|
|
||||||
|
|
||||||
### Pattern 1: Quick Planning from Task
|
|
||||||
|
|
||||||
```bash
|
|
||||||
# User has a task, needs rapid multi-perspective plan
|
|
||||||
/workflow:quick-plan-with-file -y "实现实时通知系统，支持推送和WebSocket"
|
|
||||||
# → Creates plan in ~5 minutes
|
|
||||||
# → Ready for execution
|
|
||||||
```
|
|
||||||
|
|
||||||
### Pattern 2: Convert Brainstorm to Executable Plan
|
|
||||||
|
|
||||||
```bash
|
|
||||||
# User completed brainstorm, wants to convert top idea to executable plan
|
|
||||||
/workflow:quick-plan-with-file --from brainstorm BS-notifications-2025-01-28
|
|
||||||
# → Reads synthesis.json from brainstorm
|
|
||||||
# → Generates implementation plan
|
|
||||||
# → Ready for unified-execute-with-file
|
|
||||||
```
|
|
||||||
|
|
||||||
### Pattern 3: From Analysis to Implementation
|
|
||||||
|
|
||||||
```bash
|
|
||||||
# Analysis completed, now need execution plan
|
|
||||||
/workflow:quick-plan-with-file --from analysis ANL-auth-architecture-2025-01-28
|
|
||||||
# → Reads conclusions.json from analysis
|
|
||||||
# → Generates planning with recommendations
|
|
||||||
# → Output task breakdown
|
|
||||||
```
|
|
||||||
|
|
||||||
### Pattern 4: Planning with Interactive Conflict Resolution
|
|
||||||
|
|
||||||
```bash
|
|
||||||
# Full planning with user involvement in decision-making
|
|
||||||
/workflow:quick-plan-with-file "新的支付流程集成"
|
|
||||||
# → Without -y flag
|
|
||||||
# → After conflict detection, asks user about tradeoffs
|
|
||||||
# → Generates plan based on user preferences
|
|
||||||
```
|
|
||||||
|
|
||||||
---
|
|
||||||
|
|
||||||
## Comparison with Other Workflows
|
|
||||||
|
|
||||||
| Feature | brainstorm | analyze | quick-plan | plan |
|
|
||||||
|---------|-----------|---------|-----------|------|
|
|
||||||
| **Purpose** | Ideation | Investigation | Lightweight planning | Detailed planning |
|
|
||||||
| **Multi-agent** | 3 perspectives | 2 CLI + explore | 4 concurrent agents | N/A (single) |
|
|
||||||
| **Documentation** | Extensive | Extensive | Minimal | Standard |
|
|
||||||
| **Output** | Ideas + synthesis | Conclusions | Executable tasks | IMPL_PLAN |
|
|
||||||
| **Typical Duration** | 30-60 min | 20-30 min | 5-10 min | 15-20 min |
|
|
||||||
| **User Interaction** | High (multi-round) | High (Q&A) | Low (decisions) | Medium |
|
|
||||||
|
|
||||||
---
|
|
||||||
|
|
||||||
## Error Handling
|
|
||||||
|
|
||||||
| Situation | Action |
|
|
||||||
|-----------|--------|
|
|
||||||
| Agents conflict on approach | Arbitration agent decides, document in conflicts.json |
|
|
||||||
| Missing critical files | Continue with available context, note limitations |
|
|
||||||
| Insufficient task breakdown | Ask user for planning focus areas |
|
|
||||||
| Effort estimate too high | Suggest MVP approach or phasing |
|
|
||||||
| Unclear requirements | Ask clarifying questions via AskUserQuestion |
|
|
||||||
| Agent timeout | Use last successful result, note partial analysis |
|
|
||||||
|
|
||||||
---
|
|
||||||
|
|
||||||
## Best Practices
|
|
||||||
|
|
||||||
1. **Use when**:
|
|
||||||
- You have clarity on WHAT but not HOW
|
|
||||||
- Need rapid multi-perspective planning
|
|
||||||
- Converting brainstorm/analysis into execution
|
|
||||||
- Want minimal planning overhead
|
|
||||||
|
|
||||||
2. **Avoid when**:
|
|
||||||
- Requirements are highly ambiguous (use brainstorm instead)
|
|
||||||
- Need deep investigation (use analyze instead)
|
|
||||||
- Want extensive planning document (use plan instead)
|
|
||||||
- No tech stack clarity (use analyze first)
|
|
||||||
|
|
||||||
3. **For best results**:
|
|
||||||
- Provide complete task/requirement description
|
|
||||||
- Include constraints and success criteria
|
|
||||||
- Specify preferences (speed vs quality vs risk)
|
|
||||||
- Review conflicts.json and make conscious tradeoff decisions
|
|
||||||
|
|
||||||
---
|
|
||||||
|
|
||||||
## Next Steps After Planning
|
|
||||||
|
|
||||||
### Feed to Execution
|
|
||||||
```bash
|
|
||||||
/workflow:unified-execute-with-file -p .workflow/.planning/{sessionId}/synthesis.json
|
|
||||||
```
|
|
||||||
|
|
||||||
### Detailed Planning if Needed
|
|
||||||
```bash
|
|
||||||
/workflow:plan "Based on quick-plan recommendations..."
|
|
||||||
```
|
|
||||||
|
|
||||||
### Continuous Refinement
|
|
||||||
```bash
|
|
||||||
/workflow:quick-plan-with-file --continue "{topic}" # Update plan with new constraints
|
|
||||||
```
|
|
||||||
Reference in New Issue
Block a user