Team Frontend

Auto Mode

When --yes or -y: Auto-confirm task decomposition, skip interactive validation, use defaults for scope/industry/constraints.

Team Frontend Development

Usage

$team-frontend "Build a SaaS dashboard with user management and analytics"
$team-frontend -c 3 "Create a healthcare patient portal with WCAG AA compliance"
$team-frontend -y "Implement e-commerce product listing page with dark mode"
$team-frontend --continue "fe-saas-dashboard-20260308"

Flags:

• -y, --yes: Skip all confirmations (auto mode)
• -c, --concurrency N: Max concurrent agents within each wave (default: 3)
• --continue: Resume existing session

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)

Overview

Unified frontend development team: analyze requirements, retrieve design intelligence (ui-ux-pro-max), generate design token system, architect components, implement code, and run 5-dimension quality audit. Roles: analyst, architect, developer, qa -- dynamically assigned as CSV wave tasks with dependency ordering. Supports page (4-task), feature (5-task), and system (7-task) pipeline modes.

Execution Model: Hybrid -- CSV wave pipeline (primary) + individual agent spawn (secondary)

+-------------------------------------------------------------------+
|              TEAM FRONTEND WORKFLOW                                 |
+-------------------------------------------------------------------+
|                                                                     |
|  Phase 0: Pre-Wave Interactive (Requirement Clarification)          |
|     +- Parse frontend task description                              |
|     +- Select scope (page/feature/system), industry, constraints    |
|     +- Output: refined requirements for decomposition               |
|                                                                     |
|  Phase 1: Requirement -> CSV + Classification                       |
|     +- Signal detection: keyword scan -> capability inference       |
|     +- Pipeline selection (page: 4-task, feature: 5-task, system)   |
|     +- Dependency graph from pipeline definition                    |
|     +- Classify tasks: csv-wave | interactive (exec_mode)           |
|     +- Compute dependency waves (topological sort)                  |
|     +- Generate tasks.csv with wave + exec_mode columns             |
|     +- User validates task breakdown (skip if -y)                   |
|                                                                     |
|  Phase 2: Wave Execution Engine (Extended)                          |
|     +- For each wave (1..N):                                        |
|     |   +- Execute pre-wave interactive tasks (if any)              |
|     |   +- Build wave CSV (filter csv-wave tasks for this wave)     |
|     |   +- Inject previous findings into prev_context column        |
|     |   +- spawn_agents_on_csv(wave CSV)                            |
|     |   +- Execute post-wave interactive tasks (if any)             |
|     |   +- Merge all results into master tasks.csv                  |
|     |   +- Check: any failed? -> skip dependents                    |
|     |   +- GC Loop: if QA fails, create DEV-fix + QA-recheck        |
|     +- discoveries.ndjson shared across all modes (append-only)     |
|                                                                     |
|  Phase 3: Post-Wave Interactive (Completion Action)                 |
|     +- Pipeline completion report with deliverables listing         |
|     +- Interactive completion choice (Archive/Keep/Export)           |
|     +- Final aggregation / report                                   |
|                                                                     |
|  Phase 4: Results Aggregation                                       |
|     +- Export final results.csv                                     |
|     +- Generate context.md with all findings                        |
|     +- Display summary: completed/failed/skipped per wave           |
|     +- Offer: view results | retry failed | done                    |
|                                                                     |
+-------------------------------------------------------------------+

Task Classification Rules

Each task is classified by exec_mode:

Classification Decision:

CSV Schema

tasks.csv (Master State)

id,title,description,role,pipeline_mode,scope,review_type,deps,context_from,exec_mode,wave,status,findings,artifacts_produced,qa_score,qa_verdict,error
"ANALYZE-001","Requirement analysis + design intelligence","PURPOSE: Analyze frontend requirements and retrieve design intelligence | Success: design-intelligence.json produced","analyst","feature","full","","","","csv-wave","1","pending","","","","",""
"ARCH-001","Design token system + component architecture","PURPOSE: Define design token system and component specs | Success: design-tokens.json + component specs produced","architect","feature","full","","ANALYZE-001","ANALYZE-001","csv-wave","2","pending","","","","",""
"QA-001","Architecture review","PURPOSE: Review architecture artifacts before development | Success: Architecture approved","qa","feature","full","architecture-review","ARCH-001","ARCH-001","csv-wave","3","pending","","","","",""

Columns:

Per-Wave CSV (Temporary)

Each wave generates a temporary wave-{N}.csv with extra prev_context column (csv-wave tasks only).

Agent Registry (Interactive Agents)

COMPACT PROTECTION: Agent files are execution documents. When context compression occurs, you MUST immediately Read the corresponding agent.md to reload.

Output Artifacts

Session Structure

.workflow/.csv-wave/{session-id}/
+-- tasks.csv                  # Master state (all tasks, both modes)
+-- results.csv                # Final results export
+-- discoveries.ndjson         # Shared discovery board (all agents)
+-- context.md                 # Human-readable report
+-- task-analysis.json         # Phase 1 analysis output
+-- wave-{N}.csv               # Temporary per-wave input (csv-wave only)
+-- role-instructions/         # Per-role instruction templates
|   +-- analyst.md
|   +-- architect.md
|   +-- developer.md
|   +-- qa.md
+-- artifacts/                 # All deliverables
|   +-- analysis/
|   |   +-- design-intelligence.json
|   |   +-- requirements.md
|   +-- architecture/
|   |   +-- design-tokens.json
|   |   +-- component-specs/
|   |   +-- project-structure.md
|   +-- qa/
|   |   +-- audit-001.md
|   +-- build/
+-- interactive/               # Interactive task artifacts
|   +-- {id}-result.json
+-- wisdom/                    # Cross-task knowledge
    +-- learnings.md
    +-- decisions.md

Implementation

Session Initialization

const getUtc8ISOString = () => new Date(Date.now() + 8 * 60 * 60 * 1000).toISOString()

const AUTO_YES = $ARGUMENTS.includes('--yes') || $ARGUMENTS.includes('-y')
const continueMode = $ARGUMENTS.includes('--continue')
const concurrencyMatch = $ARGUMENTS.match(/(?:--concurrency|-c)\s+(\d+)/)
const maxConcurrency = concurrencyMatch ? parseInt(concurrencyMatch[1]) : 3

const requirement = $ARGUMENTS
  .replace(/--yes|-y|--continue|--concurrency\s+\d+|-c\s+\d+/g, '')
  .trim()

const slug = requirement.toLowerCase()
  .replace(/[^a-z0-9\u4e00-\u9fa5]+/g, '-')
  .substring(0, 40)
const dateStr = getUtc8ISOString().substring(0, 10).replace(/-/g, '')
const sessionId = `fe-${slug}-${dateStr}`
const sessionFolder = `.workflow/.csv-wave/${sessionId}`

Bash(`mkdir -p ${sessionFolder}/artifacts/analysis ${sessionFolder}/artifacts/architecture/component-specs ${sessionFolder}/artifacts/qa ${sessionFolder}/artifacts/build ${sessionFolder}/role-instructions ${sessionFolder}/interactive ${sessionFolder}/wisdom`)

Write(`${sessionFolder}/discoveries.ndjson`, '')
Write(`${sessionFolder}/wisdom/learnings.md`, '# Learnings\n')
Write(`${sessionFolder}/wisdom/decisions.md`, '# Decisions\n')

Phase 0: Pre-Wave Interactive (Requirement Clarification)

Objective: Parse frontend task, clarify scope/industry/constraints, prepare for decomposition.

Workflow:

1. Parse user task description from $ARGUMENTS

2. Check for existing sessions (continue mode):

   • Scan .workflow/.csv-wave/fe-*/tasks.csv for sessions with pending tasks
   • If --continue: resume the specified or most recent session, skip to Phase 2
   • If active session found: ask user whether to resume or start new

3. Clarify scope and industry (skip if AUTO_YES):

   Scope Selection:

   Option	Pipeline	Task Count
   Single page	page	4 tasks (linear)
   Multi-component feature	feature	5 tasks (with arch review gate)
   Full frontend system	system	7 tasks (dual-track parallel)

   Industry Selection:

   Option	Strictness
   SaaS/Tech	standard
   E-commerce/Retail	standard
   Healthcare/Finance	strict (extra accessibility)
   Other	standard

   Design Constraints (multi-select): Existing design system, WCAG AA, Responsive, Dark mode

4. Record requirements: mode, scope, industry, constraints

5. Signal Detection for pipeline selection:

   Signal	Keywords	Capability
   Analysis	analyze, requirements, design intelligence	analyst
   Architecture	design tokens, component architecture, design system	architect
   Implementation	implement, build, code, develop, page, component	developer
   Quality	review, audit, quality, test, accessibility	qa

6. Complexity Scoring:

   Factor	Points
   ui-ux-pro-max integration needed	+1
   Existing design system detected	+1
   Accessibility strict mode (healthcare/finance)	+2
   Multiple tech stacks	+2
   Dark mode required	+1

   Results: 1-2 page, 3-4 feature, 5+ system. Default: feature.

Success Criteria:

• Scope, industry, constraints determined
• Pipeline mode selected (page/feature/system)

Phase 1: Requirement -> CSV + Classification

Objective: Build task dependency graph, generate tasks.csv and per-role instruction templates.

Decomposition Rules:

1. Pipeline Selection based on scope:

   Scope	Pipeline	Tasks
   page	ANALYZE-001 -> ARCH-001 -> DEV-001 -> QA-001	4
   feature	ANALYZE-001 -> ARCH-001 -> QA-001(arch) -> DEV-001 -> QA-002(code)	5
   system	ANALYZE-001 -> ARCH-001 -> QA-001(arch) -> [ARCH-002 + DEV-001] -> QA-002 -> DEV-002 -> QA-003(final)	7

2. Task Description Template: Every task description uses PURPOSE/TASK/CONTEXT/EXPECTED/CONSTRAINTS format (see dispatch.md for full templates)

3. Role Instruction Generation: Write per-role instruction templates to role-instructions/{role}.md using the base instruction template customized for each role (analyst, architect, developer, qa)

Classification Rules:

Wave Computation: Kahn's BFS topological sort with depth tracking.

// Generate per-role instruction templates
for (const role of ['analyst', 'architect', 'developer', 'qa']) {
  const instruction = generateRoleInstruction(role, sessionFolder)
  Write(`${sessionFolder}/role-instructions/${role}.md`, instruction)
}

// Generate tasks.csv from pipeline definition
const tasks = buildTasksCsv(pipelineMode, requirement, sessionFolder, industry, constraints)
Write(`${sessionFolder}/tasks.csv`, toCsv(tasks))
Write(`${sessionFolder}/task-analysis.json`, JSON.stringify(analysisResult, null, 2))

User Validation: Display task breakdown with wave + exec_mode assignment (skip if AUTO_YES).

Success Criteria:

• tasks.csv created with valid schema, wave, and exec_mode assignments
• Role instruction templates generated in role-instructions/
• task-analysis.json written
• No circular dependencies
• User approved (or AUTO_YES)

Phase 2: Wave Execution Engine (Extended)

Objective: Execute tasks wave-by-wave with hybrid mechanism support and cross-wave context propagation.

const masterCsv = Read(`${sessionFolder}/tasks.csv`)
let tasks = parseCsv(masterCsv)
const maxWave = Math.max(...tasks.map(t => t.wave))
let gcRound = 0
const MAX_GC_ROUNDS = 2

for (let wave = 1; wave <= maxWave; wave++) {
  console.log(`\nWave ${wave}/${maxWave}`)

  // 1. Separate tasks by exec_mode
  const waveTasks = tasks.filter(t => t.wave === wave && t.status === 'pending')
  const csvTasks = waveTasks.filter(t => t.exec_mode === 'csv-wave')
  const interactiveTasks = waveTasks.filter(t => t.exec_mode === 'interactive')

  // 2. Check dependencies -- skip tasks whose deps failed
  for (const task of waveTasks) {
    const depIds = (task.deps || '').split(';').filter(Boolean)
    const depStatuses = depIds.map(id => tasks.find(t => t.id === id)?.status)
    if (depStatuses.some(s => s === 'failed' || s === 'skipped')) {
      task.status = 'skipped'
      task.error = `Dependency failed: ${depIds.filter((id, i) =>
        ['failed','skipped'].includes(depStatuses[i])).join(', ')}`
    }
  }

  // 3. Execute pre-wave interactive tasks (if any)
  const preWaveInteractive = interactiveTasks.filter(t => t.status === 'pending')
  for (const task of preWaveInteractive) {
    Read(`agents/qa-gate-reviewer.md`)
    const agent = spawn_agent({
      message: `## TASK ASSIGNMENT\n\n### MANDATORY FIRST STEPS\n1. Read: agents/qa-gate-reviewer.md\n2. Read: ${sessionFolder}/discoveries.ndjson\n\nGoal: ${task.description}\nScope: ${task.title}\nSession: ${sessionFolder}\n\n### Previous Context\n${buildPrevContext(task, tasks)}`
    })
    const result = wait({ ids: [agent], timeout_ms: 600000 })
    if (result.timed_out) {
      send_input({ id: agent, message: "Please finalize and output current findings." })
      wait({ ids: [agent], timeout_ms: 120000 })
    }
    Write(`${sessionFolder}/interactive/${task.id}-result.json`, JSON.stringify({
      task_id: task.id, status: "completed", findings: parseFindings(result),
      timestamp: getUtc8ISOString()
    }))
    close_agent({ id: agent })
    task.status = 'completed'
    task.findings = parseFindings(result)
  }

  // 4. Build prev_context for csv-wave tasks
  const pendingCsvTasks = csvTasks.filter(t => t.status === 'pending')
  for (const task of pendingCsvTasks) {
    task.prev_context = buildPrevContext(task, tasks)
  }

  if (pendingCsvTasks.length > 0) {
    // 5. Write wave CSV
    Write(`${sessionFolder}/wave-${wave}.csv`, toCsv(pendingCsvTasks))

    // 6. Build instruction per role group
    const waveInstruction = buildWaveInstruction(pendingCsvTasks, sessionFolder, wave)

    // 7. Execute wave via spawn_agents_on_csv
    spawn_agents_on_csv({
      csv_path: `${sessionFolder}/wave-${wave}.csv`,
      id_column: "id",
      instruction: waveInstruction,
      max_concurrency: maxConcurrency,
      max_runtime_seconds: 900,
      output_csv_path: `${sessionFolder}/wave-${wave}-results.csv`,
      output_schema: {
        type: "object",
        properties: {
          id: { type: "string" },
          status: { type: "string", enum: ["completed", "failed"] },
          findings: { type: "string" },
          artifacts_produced: { type: "string" },
          qa_score: { type: "string" },
          qa_verdict: { type: "string" },
          error: { type: "string" }
        }
      }
    })

    // 8. Merge results into master CSV
    const results = parseCsv(Read(`${sessionFolder}/wave-${wave}-results.csv`))
    for (const r of results) {
      const t = tasks.find(t => t.id === r.id)
      if (t) Object.assign(t, r)
    }

    // 9. GC Loop Check: if a QA task returned FIX_REQUIRED
    const qaResults = results.filter(r => r.id.startsWith('QA') && r.qa_verdict === 'FIX_REQUIRED')
    for (const qr of qaResults) {
      if (gcRound < MAX_GC_ROUNDS) {
        gcRound++
        // Create DEV-fix and QA-recheck tasks, append to tasks array
        const fixId = `DEV-fix-${gcRound}`
        const recheckId = `QA-recheck-${gcRound}`
        tasks.push({
          id: fixId, title: `Fix QA issues (round ${gcRound})`,
          description: `PURPOSE: Fix issues from ${qr.id} audit | Success: All critical/high resolved\nTASK:\n- Load QA audit report\n- Fix critical/high issues\n- Re-validate\nCONTEXT:\n- Session: ${sessionFolder}\n- Upstream: ${sessionFolder}/artifacts/qa/\nEXPECTED: Fixed source files\nCONSTRAINTS: Targeted fixes only`,
          role: 'developer', pipeline_mode: tasks[0].pipeline_mode, scope: 'full',
          review_type: '', deps: qr.id, context_from: qr.id,
          exec_mode: 'csv-wave', wave: wave + 1, status: 'pending',
          findings: '', artifacts_produced: '', qa_score: '', qa_verdict: '', error: ''
        })
        tasks.push({
          id: recheckId, title: `QA recheck (round ${gcRound})`,
          description: `PURPOSE: Re-audit after fixes | Success: Score >= 8, critical == 0\nTASK:\n- Execute 5-dimension audit on fixed code\n- Focus on previously flagged issues\nCONTEXT:\n- Session: ${sessionFolder}\n- Review type: code-review\nEXPECTED: ${sessionFolder}/artifacts/qa/audit-recheck-${gcRound}.md`,
          role: 'qa', pipeline_mode: tasks[0].pipeline_mode, scope: 'full',
          review_type: 'code-review', deps: fixId, context_from: fixId,
          exec_mode: 'csv-wave', wave: wave + 2, status: 'pending',
          findings: '', artifacts_produced: '', qa_score: '', qa_verdict: '', error: ''
        })
        // Extend maxWave
      }
    }
  }

  // 10. Update master CSV
  Write(`${sessionFolder}/tasks.csv`, toCsv(tasks))

  // 11. Cleanup temp files
  Bash(`rm -f ${sessionFolder}/wave-${wave}.csv ${sessionFolder}/wave-${wave}-results.csv`)

  // 12. Display wave summary
  const completed = waveTasks.filter(t => t.status === 'completed').length
  const failed = waveTasks.filter(t => t.status === 'failed').length
  const skipped = waveTasks.filter(t => t.status === 'skipped').length
  console.log(`Wave ${wave} Complete: ${completed} completed, ${failed} failed, ${skipped} skipped`)
}

Success Criteria:

• All waves executed in order
• Both csv-wave and interactive tasks handled per wave
• Each wave's results merged into master CSV before next wave starts
• Dependent tasks skipped when predecessor failed
• GC loop (developer <-> qa) handled with max 2 rounds
• discoveries.ndjson accumulated across all waves and mechanisms

Phase 3: Post-Wave Interactive (Completion Action)

Objective: Pipeline completion report with deliverables listing and interactive completion choice.

const tasks = parseCsv(Read(`${sessionFolder}/tasks.csv`))
const completed = tasks.filter(t => t.status === 'completed')
const failed = tasks.filter(t => t.status === 'failed')

// List deliverables
const deliverables = [
  { name: "Design Intelligence", path: `${sessionFolder}/artifacts/analysis/design-intelligence.json` },
  { name: "Requirements", path: `${sessionFolder}/artifacts/analysis/requirements.md` },
  { name: "Design Tokens", path: `${sessionFolder}/artifacts/architecture/design-tokens.json` },
  { name: "Component Specs", path: `${sessionFolder}/artifacts/architecture/component-specs/` },
  { name: "Project Structure", path: `${sessionFolder}/artifacts/architecture/project-structure.md` },
  { name: "QA Audits", path: `${sessionFolder}/artifacts/qa/` }
]

console.log(`
============================================
FRONTEND TEAM COMPLETE

Pipeline: ${completed.length}/${tasks.length} tasks (${tasks[0]?.pipeline_mode} mode)
GC Rounds: ${gcRound}/${MAX_GC_ROUNDS}
Session: ${sessionFolder}

Deliverables:
${deliverables.map(d => `  - ${d.name}: ${d.path}`).join('\n')}
============================================
`)

if (!AUTO_YES) {
  AskUserQuestion({
    questions: [{
      question: "Frontend pipeline complete. What would you like to do?",
      header: "Completion",
      multiSelect: false,
      options: [
        { label: "Archive & Clean (Recommended)", description: "Archive session, output final summary" },
        { label: "Keep Active", description: "Keep session for follow-up work" },
        { label: "Export Results", description: "Export design tokens and component specs" }
      ]
    }]
  })
}

Success Criteria:

• Post-wave interactive processing complete
• User informed of deliverables and pipeline status

Phase 4: Results Aggregation

Objective: Generate final results and human-readable report.

Bash(`cp ${sessionFolder}/tasks.csv ${sessionFolder}/results.csv`)

const tasks = parseCsv(Read(`${sessionFolder}/tasks.csv`))
let contextMd = `# Frontend Development Report\n\n`
contextMd += `**Session**: ${sessionId}\n`
contextMd += `**Pipeline**: ${tasks[0]?.pipeline_mode} mode\n`
contextMd += `**Date**: ${getUtc8ISOString().substring(0, 10)}\n\n`

contextMd += `## Summary\n`
contextMd += `| Status | Count |\n|--------|-------|\n`
contextMd += `| Completed | ${tasks.filter(t => t.status === 'completed').length} |\n`
contextMd += `| Failed | ${tasks.filter(t => t.status === 'failed').length} |\n`
contextMd += `| Skipped | ${tasks.filter(t => t.status === 'skipped').length} |\n\n`

const maxWave = Math.max(...tasks.map(t => t.wave))
contextMd += `## Wave Execution\n\n`
for (let w = 1; w <= maxWave; w++) {
  const waveTasks = tasks.filter(t => t.wave === w)
  contextMd += `### Wave ${w}\n\n`
  for (const t of waveTasks) {
    const icon = t.status === 'completed' ? '[DONE]' : t.status === 'failed' ? '[FAIL]' : '[SKIP]'
    contextMd += `${icon} **${t.title}** [${t.role}]`
    if (t.qa_score) contextMd += ` Score: ${t.qa_score}/10 (${t.qa_verdict})`
    contextMd += ` ${t.findings || ''}\n\n`
  }
}

contextMd += `## QA Summary\n\n`
const qaResults = tasks.filter(t => t.role === 'qa' && t.qa_score)
for (const q of qaResults) {
  contextMd += `- **${q.id}**: Score ${q.qa_score}/10 - ${q.qa_verdict}\n`
}

Write(`${sessionFolder}/context.md`, contextMd)
console.log(`Results exported to: ${sessionFolder}/results.csv`)
console.log(`Report generated at: ${sessionFolder}/context.md`)

Success Criteria:

• results.csv exported (all tasks, both modes)
• context.md generated with QA summary
• Summary displayed to user

Shared Discovery Board Protocol

All agents (csv-wave and interactive) share a single discoveries.ndjson file for cross-task knowledge exchange.

Format: One JSON object per line (NDJSON):

{"ts":"2026-03-08T10:00:00Z","worker":"ANALYZE-001","type":"tech_stack_detected","data":{"stack":"react","framework":"nextjs","ui_lib":"shadcn"}}
{"ts":"2026-03-08T10:05:00Z","worker":"ARCH-001","type":"token_generated","data":{"category":"color","count":24,"supports_dark_mode":true}}
{"ts":"2026-03-08T10:10:00Z","worker":"DEV-001","type":"file_modified","data":{"file":"src/styles/tokens.css","change":"Generated CSS custom properties","lines_added":85}}
{"ts":"2026-03-08T10:15:00Z","worker":"QA-001","type":"issue_found","data":{"file":"src/components/Button.tsx","line":42,"severity":"high","description":"Missing cursor-pointer on button"}}

Discovery Types:

Protocol:

1. Agents MUST read discoveries.ndjson at start of execution
2. Agents MUST append relevant discoveries during execution
3. Agents MUST NOT modify or delete existing entries
4. Deduplication by {type, data.file} key

Pipeline Definitions

Page Mode (4 tasks, linear)

ANALYZE-001 --> ARCH-001 --> DEV-001 --> QA-001
[analyst]       [architect]  [developer]  [qa:code-review]
   wave 1         wave 2       wave 3       wave 4

Feature Mode (5 tasks, with architecture review gate)

ANALYZE-001 --> ARCH-001 --> QA-001 --> DEV-001 --> QA-002
[analyst]       [architect]  [qa:arch]   [developer]  [qa:code-review]
   wave 1         wave 2       wave 3       wave 4       wave 5

System Mode (7 tasks, dual-track parallel)

ANALYZE-001 --> ARCH-001 --> QA-001 --> ARCH-002 --+
[analyst]       [architect]  [qa:arch]  [architect] |
                                        DEV-001   --+--> QA-002 --> DEV-002 --> QA-003
                                        [dev:tokens]     [qa]      [dev:comp]  [qa:final]
   wave 1         wave 2       wave 3     wave 4          wave 5     wave 6     wave 7

Generator-Critic Loop (developer <-> qa)

developer (Generator) -> QA artifact -> qa (Critic)
                      <- QA feedback <-
                         (max 2 rounds)

Convergence: qa.score >= 8 && qa.critical_count === 0

Error Handling

Core Rules

1. Start Immediately: First action is session initialization, then Phase 0/1
2. Wave Order is Sacred: Never execute wave N before wave N-1 completes and results are merged
3. CSV is Source of Truth: Master tasks.csv holds all state (both csv-wave and interactive)
4. CSV First: Default to csv-wave for tasks; only use interactive when interaction pattern requires it
5. Context Propagation: prev_context built from master CSV, not from memory
6. Discovery Board is Append-Only: Never clear, modify, or recreate discoveries.ndjson
7. Skip on Failure: If a dependency failed, skip the dependent task
8. GC Loop Cap: Max 2 generator-critic rounds between developer and qa
9. Cleanup Temp Files: Remove wave-{N}.csv after results are merged
10. DO NOT STOP: Continuous execution until all waves complete or all remaining tasks are skipped

Coordinator Role Constraints (Main Agent)

CRITICAL: The coordinator (main agent executing this skill) is responsible for orchestration only, NOT implementation.

15. Coordinator Does NOT Execute Code: The main agent MUST NOT write, modify, or implement any code directly. All implementation work is delegated to spawned team agents. The coordinator only:

    • Spawns agents with task assignments
    • Waits for agent callbacks
    • Merges results and coordinates workflow
    • Manages workflow transitions between phases

16. Patient Waiting is Mandatory: Agent execution takes significant time (typically 10-30 minutes per phase, sometimes longer). The coordinator MUST:

    • Wait patiently for wait() calls to complete
    • NOT skip workflow steps due to perceived delays
    • NOT assume agents have failed just because they're taking time
    • Trust the timeout mechanisms defined in the skill

17. Use send_input for Clarification: When agents need guidance or appear stuck, the coordinator MUST:

    • Use send_input() to ask questions or provide clarification
    • NOT skip the agent or move to next phase prematurely
    • Give agents opportunity to respond before escalating
    • Example: send_input({ id: agent_id, message: "Please provide status update or clarify blockers" })

18. No Workflow Shortcuts: The coordinator MUST NOT:

    • Skip phases or stages defined in the workflow
    • Bypass required approval or review steps
    • Execute dependent tasks before prerequisites complete
    • Assume task completion without explicit agent callback
    • Make up or fabricate agent results

19. Respect Long-Running Processes: This is a complex multi-agent workflow that requires patience:

    • Total execution time may range from 30-90 minutes or longer
    • Each phase may take 10-30 minutes depending on complexity
    • The coordinator must remain active and attentive throughout the entire process
    • Do not terminate or skip steps due to time concerns