claude-flow-novice 2.14.31 → 2.14.33

package/claude-assets/agents/README-AGENT_LIFECYCLE.md

@@ -1,3 +1,11 @@
+---
+name: readme-agent-lifecycle
+description: Documentation file for agent lifecycle management and CFN loop protocols
+tools: []
+model: haiku
+type: documentation
+---
+
 # Agent Lifecycle Documentation
 
 **Version:** 2.0 (Post-Waiting Mode Removal)
@@ -52,15 +60,6 @@ This document clarifies the complete lifecycle of agents in the CFN Loop system,
 2. **Report confidence:**
    ```bash
    ./.claude/skills/redis-coordination/invoke-waiting-mode.sh report \
-
-**After reporting, exit cleanly. Do NOT enter waiting mode.**
-
-**Why This Matters:**
-- Orchestrator collects confidence/consensus scores from Redis
-- Enables adaptive agent specialization for next iteration
-- Prevents orchestrator blocking on wait $PID
-- Coordinator spawns appropriate specialist based on feedback type
-
      --task-id "$TASK_ID" \
      --agent-id "$AGENT_ID" \
      --confidence [0.0-1.0] \
@@ -106,15 +105,6 @@ redis-cli lpush "swarm:${TASK_ID}:${AGENT_ID}:done" "complete"
 ### Step 3: Report Confidence Score
 ```bash
 ./.claude/skills/redis-coordination/invoke-waiting-mode.sh report \
-
-**After reporting, exit cleanly. Do NOT enter waiting mode.**
-
-**Why This Matters:**
-- Orchestrator collects confidence/consensus scores from Redis
-- Enables adaptive agent specialization for next iteration
-- Prevents orchestrator blocking on wait $PID
-- Coordinator spawns appropriate specialist based on feedback type
-
   --task-id "$TASK_ID" \
   --agent-id "$AGENT_ID" \
   --confidence [0.0-1.0] \
@@ -165,15 +155,6 @@ redis-cli lpush "swarm:${TASK_ID}:${AGENT_ID}:done" "complete"
 ### Step 5: Report Consensus Score
 ```bash
 ./.claude/skills/redis-coordination/invoke-waiting-mode.sh report \
-
-**After reporting, exit cleanly. Do NOT enter waiting mode.**
-
-**Why This Matters:**
-- Orchestrator collects confidence/consensus scores from Redis
-- Enables adaptive agent specialization for next iteration
-- Prevents orchestrator blocking on wait $PID
-- Coordinator spawns appropriate specialist based on feedback type
-
   --task-id "$TASK_ID" \
   --agent-id "$AGENT_ID" \
   --confidence [0.0-1.0] \
@@ -299,6 +280,7 @@ CONFIDENCE=$(redis-cli get "swarm:${TASK_ID}:${AGENT_ID}:confidence")
 ### ❌ Anti-Pattern 1: Waiting Mode
 ```bash
 # WRONG - DO NOT USE:
+./.claude/skills/redis-coordination/invoke-waiting-mode.sh wait \
   --task-id "$TASK_ID" \
   --agent-id "$AGENT_ID"
 ```
@@ -346,15 +328,6 @@ redis-cli lpush "swarm:${TASK_ID}:${AGENT_ID}:done" "complete"
 
 # Step 2: Report confidence
 ./.claude/skills/redis-coordination/invoke-waiting-mode.sh report \
-
-**After reporting, exit cleanly. Do NOT enter waiting mode.**
-
-**Why This Matters:**
-- Orchestrator collects confidence/consensus scores from Redis
-- Enables adaptive agent specialization for next iteration
-- Prevents orchestrator blocking on wait $PID
-- Coordinator spawns appropriate specialist based on feedback type
-
   --task-id "$TASK_ID" \
   --agent-id "$AGENT_ID" \
   --confidence 0.95 \
@@ -519,4 +492,4 @@ swarm:{TASK_ID}:metrics:iteration_start  → List of timestamps
 
 **Version History:**
 - v1.0: Original with waiting mode
-- v2.0 (2025-10-21): Waiting mode removed, PATTERN-022 compliant
+- v2.0 (2025-10-21): Waiting mode removed, PATTERN-022 compliant

package/claude-assets/agents/README-VALIDATION.md

@@ -1,3 +1,11 @@
+---
+name: readme-validation-main
+description: Documentation file for agent profile validation tool and standards compliance
+tools: []
+model: haiku
+type: documentation
+---
+
 # Agent Profile Validation Tool
 
 A comprehensive validation script that checks agent profiles against CLAUDE.md standards and provides actionable feedback.

package/claude-assets/agents/cfn-dev-team/README.md

@@ -1,3 +1,11 @@
+---
+name: cfn-dev-team-readme
+description: Documentation file for CFN dev team agent structure and organization overview
+tools: []
+model: haiku
+type: documentation
+---
+
 # CFN Dev Team Agent Structure
 
 ## Overview

package/claude-assets/agents/cfn-dev-team/coordinators/README.md

@@ -1,3 +1,11 @@
+---
+name: coordinators-readme
+description: Documentation file for CFN dev team coordinators overview and usage patterns
+tools: []
+model: haiku
+type: documentation
+---
+
 # Coordinators
 
 Agent profiles for coordination and orchestration of multi-agent workflows.
@@ -39,4 +47,4 @@ Coordinator handles all agent spawning internally via orchestration scripts.
 ## Related Skills
 - `.claude/skills/cfn-redis-coordination/` - Multi-agent coordination patterns
 - `.claude/skills/cfn-loop-validation/` - CFN Loop mechanics
-- `.claude/skills/cfn-agent-spawning/` - CLI spawning patterns
+- `.claude/skills/cfn-agent-spawning/` - CLI spawning patterns

package/claude-assets/agents/cfn-dev-team/coordinators/cfn-v3-coordinator.md

@@ -16,18 +16,16 @@ You coordinate CFN Loop v3 execution with Redis-based agent orchestration and CL
 
 Orchestrate CFN Loop v3 execution using Redis coordination for CLI agent spawning, context management, and consensus collection.
 
-## Mode-Specific Execution
+## CLI Mode Execution (Coordinator is ALWAYS CLI Mode)
 
-**CLI Mode (Production):**
+**IMPORTANT:** This coordinator agent is **ALWAYS CLI mode**. Task mode coordination happens directly in Main Chat, not via coordinator agents.
+
+When spawned via CLI (`npx claude-flow-novice agent-spawn`), implement full Redis coordination:
 - Use Redis coordination for agent spawning
 - Store context in Redis for swarm recovery
 - Collect confidence scores via Redis signals
 - Use background execution with monitoring
-
-**Task Mode (Debugging):**
-- Direct Task() spawning (no Redis needed)
-- Main Chat handles coordination
-- Simple JSON response only
+- **ALWAYS invoke the orchestrator** - never handle tasks directly
 
 ## Redis Coordination Protocols
 
@@ -494,11 +492,75 @@ Extract deliverables from task description:
    - Estimate iteration count
    - Provide reasoning for choices
 
+## Execution Steps (CLI Mode Only)
+
+**CRITICAL:** This coordinator is ALWAYS CLI mode. There is no Task Mode execution path.
+
+### Step 1: Task Classification (REQUIRED)
+```bash
+# Classify task type (use hardcoded defaults if script fails)
+TASK_TYPE="infrastructure"  # Default fallback
+if [[ -f ".claude/skills/task-classifier/classify-task.sh" ]]; then
+  CLASSIFIED_TYPE=$(bash .claude/skills/task-classifier/classify-task.sh "$TASK_DESCRIPTION" 2>/dev/null || echo "")
+  [[ -n "$CLASSIFIED_TYPE" ]] && TASK_TYPE="$CLASSIFIED_TYPE"
+fi
+```
+
+### Step 2: Agent Selection with Fallback (REQUIRED)
+```bash
+# Select agents with hardcoded fallbacks (never fail)
+LOOP3_AGENTS="terraform-engineer,devops-engineer"  # Infrastructure default
+LOOP2_AGENTS="security-auditor,compliance-checker,cost-optimizer"  # Validation default
+PRODUCT_OWNER="product-owner"
+
+# Try dynamic selection but fall back to defaults
+if [[ -f ".claude/skills/cfn-agent-selector/select-agents.sh" ]]; then
+  SELECTED_AGENTS=$(bash .claude/skills/cfn-agent-selector/select-agents.sh --task-type "$TASK_TYPE" --description "$TASK_DESCRIPTION" 2>/dev/null || echo "")
+  if [[ -n "$SELECTED_AGENTS" ]]; then
+    PARSED_AGENTS=$(echo "$SELECTED_AGENTS" | jq -r '.loop3 // empty | join(",")')
+    [[ -n "$PARSED_AGENTS" ]] && LOOP3_AGENTS="$PARSED_AGENTS"
+
+    PARSED_AGENTS=$(echo "$SELECTED_AGENTS" | jq -r '.loop2 // empty | join(",")')
+    [[ -n "$PARSED_AGENTS" ]] && LOOP2_AGENTS="$PARSED_AGENTS"
+  fi
+fi
+```
+
+### Step 3: INVOKE ORCHESTRATOR (MANDATORY - NOT OPTIONAL)
+
+**CRITICAL:** You MUST invoke orchestrator by iteration 3. DO NOT complete tasks directly.
+
+```bash
+# ALWAYS invoke orchestrator - this is your ONLY responsibility
+./.claude/skills/cfn-loop-orchestration/orchestrate.sh \
+  --task-id "$TASK_ID" \
+  --mode "standard" \
+  --loop3-agents "$LOOP3_AGENTS" \
+  --loop2-agents "$LOOP2_AGENTS" \
+  --product-owner "$PRODUCT_OWNER" \
+  --max-iterations 10 \
+  --success-criteria '{"deliverables":[],"acceptanceCriteria":["Implementation complete"]}'
+
+# The orchestrator handles ALL CFN Loop execution including:
+# - Loop 3 agent spawning and iteration
+# - Loop 2 validator spawning and consensus
+# - Product Owner decision (PROCEED/ITERATE/ABORT)
+# - Git commit and push (on PROCEED)
+# - Sprint summary generation
+
+# Your job is complete after invoking this command - DO NOT wait for results
+```
+
+**EXECUTION GUARANTEE:**
+- If steps 1-2 fail, use hardcoded defaults and proceed to step 3
+- **Never exit without invoking orchestrator**
+- **Orchestrator invocation MUST happen by iteration 3**
+- This coordinator's ONLY job is to configure and invoke the orchestrator
+
 ## Success Metrics
 
 - Agent selections match domain expertise
 - Validation criteria cover all critical requirements
 - Deliverable list is comprehensive
 - Confidence score ≥ 0.85 in analysis quality
-
-Provide structured output with confidence score based on analysis completeness and agent selection appropriateness.
+- **CRITICAL: Orchestrator invoked successfully**

package/claude-assets/agents/cfn-dev-team/developers/README.md

@@ -1,3 +1,11 @@
+---
+name: developers-readme
+description: Documentation file for CFN dev team developers overview and implementation standards
+tools: []
+model: haiku
+type: documentation
+---
+
 # Developers
 
 Implementation agents focused on building features and components.
@@ -66,4 +74,4 @@ Developers MUST run after every file edit:
 ./.claude/hooks/cfn-invoke-post-edit.sh "$EDITED_FILE" --agent-id "$AGENT_ID"
 ```
 
-This ensures code quality, test coverage, and security validation.
+This ensures code quality, test coverage, and security validation.

package/claude-assets/agents/cfn-dev-team/documentation/README-VALIDATION.md

@@ -1,3 +1,11 @@
+---
+name: readme-validation
+description: Documentation file for agent profile validation tool usage and standards
+tools: []
+model: haiku
+type: documentation
+---
+
 # Agent Profile Validation Tool
 
 A comprehensive validation script that checks agent profiles against CLAUDE.md standards and provides actionable feedback.

package/claude-assets/agents/cfn-dev-team/documentation/agent-type-guidelines.md

@@ -1,3 +1,13 @@
+---
+name: agent-type-guidelines
+description: MUST BE USED when creating agent type guidelines and documentation. Use PROACTIVELY for agent architecture, type definitions, development standards. Keywords - agent, guidelines, types, architecture, development
+tools: [Read, Write, Edit, Bash, Grep, Glob, TodoWrite]
+model: sonnet
+type: specialist
+acl_level: 2
+capabilities: [agent-documentation, type-guidelines, architecture-standards]
+---
+
 # Agent Type Guidelines
 
 **Version:** 2.0.0

package/claude-assets/agents/cfn-dev-team/reviewers/README.md

@@ -1,3 +1,11 @@
+---
+name: reviewers-readme
+description: Documentation file for CFN dev team reviewers overview and code review standards
+tools: []
+model: haiku
+type: documentation
+---
+
 # Reviewers
 
 Code review and validation agents focused on quality assurance.
@@ -55,4 +63,4 @@ Structured feedback with:
 - Critical issues (blocking)
 - Warnings (should fix)
 - Suggestions (improvements)
-- Specific file/line references
+- Specific file/line references

package/claude-assets/agents/cfn-dev-team/reviewers/quality/quality-metrics.md

@@ -1,3 +1,13 @@
+---
+name: quality-metrics
+description: MUST BE USED when measuring agent effectiveness and code quality validation. Use PROACTIVELY for metrics analysis, quality assessment, performance validation. Keywords - quality, metrics, validation, measurement, effectiveness
+tools: [Read, Write, Edit, Bash, Grep, Glob, TodoWrite]
+model: haiku
+type: validator
+acl_level: 2
+capabilities: [quality-metrics, performance-analysis, validation]
+---
+
 # Quality Metrics & Validation
 
 **Version:** 2.0.0

package/claude-assets/agents/cfn-dev-team/test-agent.md

@@ -1,3 +1,13 @@
+---
+name: test-agent
+description: MUST BE USED when documenting epic creator deletion issues and bug fixes. Use PROACTIVELY for issue tracking, bug documentation, resolution summaries. Keywords - test, bug, issue, documentation, epic, deletion
+tools: [Read, Write, Edit, Bash, Grep, Glob, TodoWrite]
+model: haiku
+type: specialist
+acl_level: 1
+capabilities: [bug-documentation, issue-tracking, test-summary]
+---
+
 # Epic Creator Deletion Issue - Complete Summary
 
 **Date:** 2025-11-03

package/claude-assets/agents/cfn-dev-team/testers/README.md

@@ -1,3 +1,11 @@
+---
+name: testers-readme
+description: Documentation file for CFN dev team testers overview and testing standards
+tools: []
+model: haiku
+type: documentation
+---
+
 # Testers
 
 Testing and validation specialists ensuring code quality and correctness.
@@ -91,4 +99,4 @@ Structured report with:
 - Test execution summary
 - Coverage analysis
 - Failed test details
-- Recommendations for improvement
+- Recommendations for improvement

package/claude-assets/agents/csuite/cto-agent.md

@@ -1,3 +1,13 @@
+---
+name: cto-agent
+description: MUST BE USED when making strategic technical decisions and architectural oversight. Use PROACTIVELY for technical roadmap planning, architecture reviews, strategic guidance. Keywords - CTO, technical strategy, architecture, leadership, oversight
+tools: [Read, Write, Edit, Bash, Grep, Glob, TodoWrite]
+model: sonnet
+type: coordinator
+acl_level: 4
+capabilities: [technical-strategy, architecture-oversight, leadership]
+---
+
 # CTO Agent - Dr. Tech
 
 ## Role Identity

package/claude-assets/agents/custom/cfn-system-expert.md

@@ -256,4 +256,131 @@ redis-cli lpush "swarm:${TASK_ID}:${AGENT_ID}:done" "complete"
 - `.claude/skills/cfn-redis-coordination/SKILL.md`: Coordination patterns
 - `.claude/commands/cfn/CFN_COORDINATOR_PARAMETERS.md`: Parameter specifications
 - `planning/cfn-v3/DUAL_MODE_IMPLEMENTATION.md`: Architecture details
-- Planning documents: Sprint lessons and adaptive context insights
+- Planning documents: Sprint lessons and adaptive context insights
+
+## CFN Implementation Patterns & Lessons Learned
+
+### Core Architecture Patterns
+
+#### STRAT-007: Background Execution Strategy
+- **Insight**: Use background execution with coordination monitoring for long-running orchestration workflows (>10 minutes). Bash tool has hard 10-minute timeout that cannot be extended.
+- **Application**: Essential for CFN Loop orchestration that exceeds standard timeout limits.
+
+#### PATTERN-008: Product Owner Decision Flow
+- **Insight**: Implement explicit Product Owner decision flow after Loop 2 consensus to prevent validator scope creep and enforce strategic boundaries.
+- **Application**: Core CFN Loop governance pattern ensuring proper decision hierarchy.
+
+#### PATTERN-010: Parallel Execution with Temp Files
+- **Insight**: Use background processes with temporary files to eliminate race conditions in parallel agent coordination. Pattern: spawn all agents in background with `(skill-execution > /tmp/output-file) &`, collect PIDs, use `wait` to synchronize, read results from temp files after completion.
+- **Application**: Optimizes CFN Loop parallel agent execution while maintaining coordination integrity.
+
+### Context Management Patterns
+
+#### STRAT-021: Standardized Context Extraction Templates
+- **Insight**: Use standardized context extraction templates in coordinators. Template structure: epicGoal (1-2 sentences), inScope (list), outOfScope (list), deliverables (file paths), directory (creation path), acceptanceCriteria (measurable requirements).
+- **Application**: Prevents minimal context that causes wrong deliverables in CFN execution.
+
+#### PATTERN-020: Multi-Layer Context Injection
+- **Context**: Coordinator → Orchestrator → Agent Flow
+- **Insight**: When implementing multi-layer coordination, ensure context flows through ALL layers. Breaking this chain causes 'consensus on vapor' (high confidence, zero deliverables).
+- **Application**: Critical for CFN Loop context integrity across coordinator-orchestrator-agent chain.
+
+#### PATTERN-021: Context Validation Pipeline
+- **Context**: Multi-Layer Coordination
+- **Insight**: Design context validation pipeline with checkpoints at each layer. Pattern: (1) Coordinator validates extracted context, (2) Orchestrator validates persistence retrieval, (3) Agents validate received context.
+- **Application**: Fail-fast prevents cascading context loss in CFN workflows.
+
+#### PATTERN-022: Agent Lifecycle - Clean Exit Protocol
+- **Insight**: Agents must exit cleanly after reporting confidence. Pattern: (1) Signal done via coordination protocol, (2) Report confidence score, (3) Exit immediately.
+- **Application**: Enables adaptive agent specialization and prevents orchestrator blocking.
+
+### Sprint Execution Patterns
+
+#### Sprint Context Structure
+```json
+{
+  "sprint_name": "P1 Coordinator Monitoring",
+  "sprint_num": 1,
+  "total_sprints": 7,
+  "deliverables": [
+    "test-p1-monitoring.sh",
+    "docs/P1_MONITORING_RESULTS.md"
+  ],
+  "in_scope": [
+    "P1 coordinator monitoring validation",
+    "Test timeout mechanisms",
+    "Logging verification"
+  ],
+  "out_of_scope": [
+    "P2-P7 monitoring",
+    "Cross-priority integration",
+    "Epic-level summary"
+  ],
+  "directory": "/mnt/c/Users/masha/Documents/claude-flow-novice/tests/p1"
+}
+```
+
+#### Sprint Execution Tool: `execute-sprint-task.sh`
+- **Purpose**: Enables sprint-aware agent execution with focused context injection
+- **Usage**: Decompose large epics into focused sprints with clear deliverables and scope boundaries
+- **Key Principles**: Focused scope, incremental progress, precise deliverables, context specificity
+
+### CFN v3 Advanced Patterns
+
+#### STRAT-026: Persistence Context Storage Over CLI Parameters
+- **Insight**: Use persistence layer for complex JSON context storage instead of CLI parameters. Eliminates shell escaping issues, enables swarm recovery, and provides single source of truth.
+- **Application**: Foundation for CFN v3 dual-mode architecture.
+
+#### PATTERN-023: Dual-Mode Architecture Pattern
+- **Insight**: Implement dual execution modes (optimized vs simplified) sharing core logic. CLI mode for production (cost-optimized), Task mode for debugging (full visibility).
+- **Application**: Enables flexible CFN execution based on user needs.
+
+#### PATTERN-024: Swarm Recovery via Persistence
+- **Insight**: Store swarm state in persistence layer with TTL to enable crash recovery. Agents can resume from last known state using task_id as recovery key.
+- **Application**: Critical for long-running CFN Loops and system reliability.
+
+#### STRAT-028: Modular Skill Architecture
+- **Insight**: Decompose complex systems into independent skills (20 skills in CFN v3). Enables reuse, testing isolation, and incremental enhancement.
+- **Application**: Core architectural principle for CFN system maintainability.
+
+### CFN Anti-Patterns & Prevention
+
+#### ANTI-020: Context Storage Without Injection
+- **Context**: Coordination Systems
+- **Insight**: Avoid storing context in persistence layer without retrieving and injecting it into agent prompts.
+- **Prevention**: Always validate context injection chain: storage → retrieval → agent delivery.
+
+#### ANTI-021: Generic Context When Specifics Exist
+- **Context**: Agent Spawning
+- **Insight**: Never pass generic iteration-level context when task-specific deliverables exist.
+- **Prevention**: Always inject complete deliverables, directory paths, and acceptance criteria.
+
+#### ANTI-023: Task-Spawned Validators Without Completion Protocol
+- **Context**: Loop 2 Validation (Task Mode)
+- **Insight**: Main Chat spawns validators via Task() without clear scope boundaries or mode-aware completion protocols.
+- **Prevention**: Task-spawned validators should return structured output only - no coordination signals.
+
+### CFN Testing & Debugging
+
+#### EDGE-020: Comparative Agent Spawn Testing
+- **Context**: Context Debugging
+- **Insight**: When debugging context issues, use comparative agent spawn testing. Manual spawn with explicit context vs CFN Loop with generic context reveals injection failures.
+- **Application**: Primary debugging technique for context-related CFN issues.
+
+### Universal Anti-Patterns (CFN Relevance)
+
+#### ANTI-004: Regex Validation Anti-Pattern
+- **Insight**: Avoid simplistic regex matching for agent validation. Pattern `[[ $AGENTS =~ $AGENTS ]]` always returns true.
+- **CFN Relevance**: Critical for agent validation logic in CFN coordination.
+
+#### PATTERN-025: Comprehensive File Validation
+- **Insight**: Implement multi-stage validation including file type, permissions, size constraints, and content integrity.
+- **CFN Relevance**: Essential for CFN deliverable verification and security.
+
+#### PATTERN-026: Shell Strict Mode
+- **Insight**: Enable shell strict mode using `set -euo pipefail` for robust and predictable shell scripts.
+- **CFN Relevance**: Critical reliability pattern for all CFN skill scripts.
+
+#### PATTERN-028: Process Group Management
+- **Insight**: Implement comprehensive process group management for clean termination and resource cleanup.
+- **CFN Relevance**: Essential for CFN background process orchestration and preventing resource leaks.

package/claude-assets/agents/docker-coordinators/cfn-docker-v3-coordinator.md

@@ -1,7 +1,11 @@
 ---
-description: "CFN Docker v3 Coordinator - Container-based agent orchestration with skill-based MCP isolation"
+name: cfn-docker-v3-coordinator
+description: MUST BE USED when orchestrating container-based CFN Loop execution with MCP isolation. Use PROACTIVELY for docker-based agent coordination, skill selection, resource management. Keywords - docker, containers, orchestration, MCP, isolation
 argument-hint: "[task-description] --mode=mvp|standard|enterprise --memory-limit=1g --docker-network=mcp-network"
 allowed-tools: ["Bash", "Read", "Write", "Edit", "Grep", "Glob", "TodoWrite", "Task"]
+model: sonnet
+type: coordinator
+acl_level: 3
 ---
 
 # CFN Docker V3 Coordinator

package/claude-assets/agents/docker-team/csuite/c-suite-template.md

@@ -1,8 +1,12 @@
 ---
-description: C-Suite executive personas for strategic decision-making in organizational architecture
+name: c-suite-template
+description: MUST BE USED when creating C-Suite executive personas for strategic decision-making. Use PROACTIVELY for organizational architecture, executive strategy, decision authority. Keywords - c-suite, executive, strategy, decision-making, leadership
 tools: [Read, Write, TodoWrite, Bash]
 priority: medium
 tags: [csuite, executive, strategy, decision-making, cmo, cfo, coo, ceo]
+model: sonnet
+type: specialist
+acl_level: 4
 ---
 
 # C-Suite Agent Template

package/claude-assets/agents/docker-team/infrastructure/team-coordinator-template.md

@@ -1,8 +1,12 @@
 ---
-description: Generate team-specific coordinator configurations for Docker organizational architecture
+name: team-coordinator-template
+description: MUST BE USED when generating team-specific coordinator configurations for Docker architecture. Use PROACTIVELY for infrastructure templating, team configuration, Docker services. Keywords - docker, infrastructure, templating, coordinator, team-config
 tools: [Read, Write, Edit, Bash, Grep, Glob]
 priority: high
 tags: [docker, infrastructure, templating, coordinator, team-config]
+model: haiku
+type: specialist
+acl_level: 3
 ---
 
 # Team Coordinator Template Agent

package/claude-assets/agents/marketing_hybrid/cost_tracker.md

@@ -1,3 +1,13 @@
+---
+name: cost-tracker
+description: MUST BE USED when monitoring Z.ai worker costs during pilot testing. Use PROACTIVELY for cost tracking, budget monitoring, expense reporting. Keywords - cost, tracking, monitoring, Z.ai, budget, expenses
+tools: [Read, Write, Edit, Bash, Grep, Glob, TodoWrite]
+model: haiku
+type: specialist
+acl_level: 2
+capabilities: [cost-monitoring, budget-tracking, expense-reporting]
+---
+
 # Cost Tracking Agent
 ## Role
 Monitor Z.ai worker costs during 48-hour pilot

package/claude-assets/agents/marketing_hybrid/docker_deployer.md

@@ -1,3 +1,13 @@
+---
+name: docker-deployer
+description: MUST BE USED when deploying marketing coordinator services via Docker. Use PROACTIVELY for container deployment, service configuration, docker-compose management. Keywords - docker, deploy, containers, docker-compose, services
+tools: [Read, Write, Edit, Bash, Grep, Glob, TodoWrite]
+model: haiku
+type: specialist
+acl_level: 2
+capabilities: [docker-deployment, container-management, service-configuration]
+---
+
 # Docker Deployer Agent
 ## Role
 Deploy marketing coordinator using docker-compose.hybrid.yml

package/claude-assets/agents/marketing_hybrid/zai_worker_spawner.md

@@ -1,3 +1,13 @@
+---
+name: zai-worker-spawner
+description: MUST BE USED when spawning Z.ai workers for marketing pilot testing. Use PROACTIVELY for worker orchestration, API configuration, service scaling. Keywords - Z.ai, worker, spawner, scaling, orchestration
+tools: [Read, Write, Edit, Bash, Grep, Glob, TodoWrite]
+model: haiku
+type: specialist
+acl_level: 2
+capabilities: [worker-spawning, service-orchestration, api-configuration]
+---
+
 # Z.ai Worker Spawner Agent
 ## Role
 Spawn 3 Z.ai workers for marketing pilot

package/claude-assets/commands/CFN_LOOP_TASK_MODE.md

@@ -467,7 +467,7 @@ Complex/Enterprise (>5 files, >500 LOC): +code-analyzer
 ### Complexity Analysis CLI
 ```bash
 ./.claude/skills/task-complexity/analyze.sh --task "$DESC" --files "$LIST"
-# Output: {"score": 85, "category": "standard", "validators": ["reviewer","tester","architect","security-specialist"], "threshold": 0.90}
+# Output: {"score": 85, "category": "standard", "validators": ["reviewer","tester","system-architect","security-specialist"], "threshold": 0.90}
 ```
 
 ---

package/claude-assets/hooks/cfn-post-execution/memory-cleanup.sh

@@ -0,0 +1,20 @@
+#!/bin/bash
+
+# CFN Post-Execution Memory Cleanup Hook
+# Automatically cleans up memory resources after CFN operations
+
+set -euo pipefail
+
+# Hook configuration
+HOOK_SCRIPT_PATH="$(dirname "$0")/memory-cleanup.sh"
+MEMORY_MANAGEMENT_SKILL="$(dirname "$0")/../../skills/cfn-memory-management/cleanup-memory.sh"
+
+# Use the memory management skill if available, otherwise use local fallback
+if [[ -f "$MEMORY_MANAGEMENT_SKILL" ]]; then
+    exec "$MEMORY_MANAGEMENT_SKILL" "$@"
+elif [[ -f "$HOOK_SCRIPT_PATH" ]]; then
+    exec "$HOOK_SCRIPT_PATH" "$@"
+else
+    echo "WARNING: Memory cleanup scripts not found, skipping cleanup"
+    exit 0
+fi

package/claude-assets/hooks/cfn-pre-execution/memory-check.sh

@@ -0,0 +1,20 @@
+#!/bin/bash
+
+# CFN Pre-Execution Memory Check Hook
+# Automatically checks memory availability before CFN operations
+
+set -euo pipefail
+
+# Hook configuration
+HOOK_SCRIPT_PATH="$(dirname "$0")/memory-check.sh"
+MEMORY_MANAGEMENT_SKILL="$(dirname "$0")/../../skills/cfn-memory-management/check-memory.sh"
+
+# Use the memory management skill if available, otherwise use local fallback
+if [[ -f "$MEMORY_MANAGEMENT_SKILL" ]]; then
+    exec "$MEMORY_MANAGEMENT_SKILL" "$@"
+elif [[ -f "$HOOK_SCRIPT_PATH" ]]; then
+    exec "$HOOK_SCRIPT_PATH" "$@"
+else
+    echo "WARNING: Memory check scripts not found, proceeding without memory validation"
+    exit 0
+fi