Skip to main content Início Criadores ajbcoding claude-skill-eval moai-alfred-session-state
moai-alfred-session-state Enterprise session state management, token budget optimization, runtime tracking, session handoff protocols, context continuity for Claude Sonnet 4.5 and Haiku 4.5 with context awareness features
Ir para a instalação Skills Marketplace Descubra e explore skills de IA criadas pela comunidade.
Instalar com Codex ou Claude Copie este prompt, cole no Codex, Claude ou outro assistente e deixe que ele revise a página da skill e instale para você.
Mostrar detalhes do prompt Um comando direto ignora o prompt de revisão. Verifique a origem antes de executá-lo.
npx skills add https://github.com/AJBcoding/claude-skill-eval --skill moai-alfred-session-stateO comando permanece em uma só linha. Role horizontalmente para revisá-lo antes de copiar.
Prefere uma cópia local? Baixe os arquivos disponíveis atualmente no SkillsMP.
Baixar Zip Baixando... Mais deste repositório Enterprise database architecture specialist with PostgreSQL 17, MySQL 8.4 LTS,
MongoDB 8.0, Redis 7.4 expertise. Master connection pooling, query optimization,
caching strategies, and database DevOps automation. Build scalable, resilient
database systems with comprehensive monitoring and disaster recovery.
Enterprise Frontend Development with AI-powered modern architecture, Context7 integration, and intelligent component orchestration for scalable user interfaces
Enterprise-grade security expertise with production-ready patterns for OWASP Top 10 2021, zero-trust architecture, threat modeling (STRIDE, PASTA), secure SDLC, DevSecOps automation, cloud security, cryptography, identity & access management, and compliance frameworks (SOC 2, ISO 27001, GDPR, CCPA).
Ocupações relacionadas SOC
Baseado na classificação ocupacional SOC
Explorador de arquivos
3 arquivos name moai-alfred-session-state version 4.0.0 tier Alfred created 2025-11-05T00:00:00.000Z updated 2025-11-12T00:00:00.000Z status stable description Enterprise session state management, token budget optimization, runtime tracking, session handoff protocols, context continuity for Claude Sonnet 4.5 and Haiku 4.5 with context awareness features keywords ["session","state","context-window","token-budget","handoff","continuity","context-awareness"] allowed-tools ["Read","Bash","TodoWrite"]
Alfred Session State Management Skill (Enterprise v4.0.0)
Skill Metadata
Field Value Skill Name moai-alfred-session-state Version 4.0.0 (Enterprise) Updated 2025-11-12 Status Active Tier Alfred Supported Models Claude Sonnet 4.5, Claude Haiku 4.5 Context Window 200K tokens (Sonnet/Haiku), 500K tokens (Enterprise), 1M tokens (beta) Key Features Context Awareness, Token Budget Tracking, Session Persistence, Adaptive Recovery
What It Does
Provides enterprise-grade session state management for extended workflows, token budget optimization, runtime tracking, and handoff protocols to maintain context continuity across Alfred workflows and session boundaries.
Enterprise v4.0.0 Capabilities :
✅ Context-aware token budget management (November 2025 Claude API features)
✅ Session persistence with automatic history loading
✅ Session forking for parallel exploration
✅ Incremental multi-pack index optimization (Git 2.47+ integration)
✅ Context continuity across handoffs with state snapshots
✅ Progressive disclosure for memory efficiency
✅ Adaptive recovery checkpoints
✅ Multi-agent coordination protocols
✅ Memory file state synchronization
✅ Token budget awareness callbacks (Sonnet/Haiku 4.5 feature)
When to Use
Automatic triggers :
Session start/end events
Long-running task execution (>10 minutes)
Multi-agent handoffs
Context window approaching limits
Model switches (Haiku ↔ Sonnet)
Workflow phase transitions
Manual reference :
Session state debugging and recovery
Token budget optimization strategies
Handoff protocol design
Context continuity planning
Multi-session workflow design
Token Budget Management (November 2025)
Context Awareness Feature Claude Sonnet 4.5 and Haiku 4.5 feature built-in context awareness , enabling these models to:
Track remaining context window ("token budget") throughout conversation
Understand current position within 200K token limit (Sonnet/Haiku)
Execute adaptive strategies based on available tokens
Automatically manage context without manual intervention
Key Advantage : Models now self-regulate context usage in real time.
Token Budget Optimization Framework Token Allocation Strategy (200K Sonnet context):
├── System Prompt & Instructions: ~15K tokens (7.5%)
│ ├── CLAUDE.md: ~8K
│ ├── Command definitions: ~4K
│ └── Skill metadata: ~3K
├── Active Conversation: ~80K tokens (40%)
│ ├── Recent messages: ~50K
│ ├── Context cache: ~20K
│ └── Active references: ~10K
├── Reference Context (Progressive Disclosure): ~50K (25%)
│ ├── Project structure: ~15K
│ ├── Related Skills: ~20K
│ └── Tool definitions: ~15K
└── Reserve (Emergency Recovery): ~55K tokens (27.5%)
├── Session state snapshot: ~10K
├── TAGs and cross-references: ~15K
├── Error recovery context: ~20K
└── Free buffer: ~10K
Optimization Techniques (v4.0.0) Technique 1: Progressive Summarization
Step 1: Original context (50K tokens)
Step 2: Compress to summary (15K tokens)
Step 3: Add pointers to original → 35K tokens saved
Step 4: Carry forward summary only across handoffs
Technique 2: Context Tagging with Unique Identifiers
❌ Bad (high token cost):
"The user configuration from the previous 20 messages..."
✅ Good (efficient reference):
Technique 3: Structured Context Architecture
├── Critical Context (ALWAYS keep)
│ ├── Current task objectives
│ ├── User preferences & expertise level
│ └── Active constraints
├── Supporting Context (keep if space allows)
│ ├── Related history
│ ├── Reference documentation
│ └── Tool availability
└── Temporary Context (discard when not needed)
├── Raw tool outputs
├── Intermediate calculations
└── Debug information
Technique 4: MCP Server Context Budget
Technique 5: Task-Based Session Management
Strategy: Start new conversation for distinct tasks
Benefits:
- Fresh 200K token budget per task
- Eliminates stale context accumulation
- Enables parallel session forking
- Improves recovery speed
Implementation:
1. Complete current task in Session A
2. Save session snapshot to .moai/sessions/
3. Start Session B for new task with fresh context
4. Resume Session A later if needed via session ID
Session State Architecture (Enterprise v4.0.0)
State Layers Session State Stack (Enterprise v4.0.0):
├── L1: Context-Aware Layer (Claude 4.5+ feature)
│ ├── Token budget tracking
│ ├── Context window position
│ ├── Auto-summarization triggers
│ └── Model-specific optimizations
├── L2: Active Context (current task, variables, scope)
├── L3: Session History (recent actions, decisions)
├── L4: Project State (SPEC progress, milestones)
├── L5: User Context (preferences, language, expertise)
└── L6: System State (tools, permissions, environment)
Session Creation & Persistence Agent SDK Session Management (November 2025 API):
{
"session_id" : "sess_uuid_v4" ,
"model" : "claude-sonnet-4-5-20250929" ,
"created_at" : "2025-11-12T10:30:00Z" ,
"context_window" : {
"total" : 200000 ,
"used" : 85000 ,
"available" : 115000 ,
"position_percent" : 42.5
} ,
"persistence" : {
"auto_load_history" : true ,
"context_preservation" : "critical_only" ,
"cache_enabled" : true } ,
"forking" : {
"enabled" : true ,
"fork_session_id" : "sess_fork_uuid" ,
"checkpoint_timestamp" : "2025-11-12T10:30:00Z"
}
}
Session Resumption Pattern
session_id = extract_session_id(response)
save_session_checkpoint({
'session_id' : session_id,
'timestamp' : now(),
'model' : 'claude-sonnet-4-5' ,
'context_state' : current_context_snapshot()
})
response = claude.messages.create(
model="claude-sonnet-4-5-20250929" ,
resume=session_id,
messages=[new_message]
)
response = claude.messages.create(
model="claude-sonnet-4-5-20250929" ,
fork_session=session_id,
messages=[alternative_message]
)
Runtime State Tracking
Task State Machine (Enterprise v4.0.0) Workflow State Transitions:
pending → in_progress → blocked (waiting) → completed/failed
↓ ↓
[monitor token budget] [save checkpoint]
[track elapsed time] [update history]
[check for recovery] [archive state]
pending - Queued but not startedin_progress - Currently executing (monitor tokens)blocked - Waiting for dependencies or inputtoken_warning - Approaching context limit (v4.0.0)context_switch - Model change or session forkcompleted - Finished successfullyfailed - Error occurred, initiating recoveryrecovered - Resumed from checkpoint
Token Budget Callbacks (Haiku/Sonnet 4.5 Feature) def token_budget_callback (context ):
"""
Called automatically when token budget changes.
Model provides real-time context awareness.
"""
remaining_tokens = context.available_tokens
used_percent = context.token_usage_percent
if used_percent > 85 :
compress_context_window()
archive_old_context()
elif used_percent > 75 :
defer_non_critical_context()
elif used_percent > 60 :
track_context_growth()
Session Handoff Protocols
Inter-Agent Handoff Package (Enterprise v4.0.0) {
"handoff_id" : "uuid-v4" ,
"timestamp" : "2025-11-12T10:30:00Z" ,
"from_agent" : "spec-builder" ,
"to_agent" : "tdd-implementer" ,
"session_context" : {
"session_id" : "sess_uuid" ,
"model" : "claude-sonnet-4-5-20250929" ,
"context_position" : 42.5 ,
"available_tokens" : 115000 ,
"user_language" : "ko" ,
"expertise_level" : "intermediate" ,
"current_project" : "MoAI-ADK"
} ,
"task_context" : {
"spec_id" : "SPEC-001" ,
"current_phase" : "implementation" ,
"completed_steps" : [ "spec_complete" , "architecture_defined" ] ,
"next_step" : "write_tests" ,
"constraints" : [ "must_use_pytest" , "coverage_85" ]
} ,
"context_snapshot" : {
"critical_context" : "...compressed..." ,
"session_checkpoints" : [ ...] ,
"active_todos" : [ ...] ,
"token_budget_strategy" : "progressive_summarization"
} ,
"recovery_info" : {
"last_checkpoint" : "2025-11-12T10:25:00Z" ,
"recovery_tokens_reserved" : 55000 ,
"session_fork_available" : true }
}
Handoff Validation (Enterprise v4.0.0) def validate_handoff (handoff_package ):
"""Enterprise validation with token budget check"""
required_fields = [
'handoff_id' , 'from_agent' , 'to_agent' ,
'session_context' , 'task_context' , 'context_snapshot'
]
for field in required_fields:
if field not in handoff_package:
raise HandoffError(f"Missing required field: {field} " )
context = handoff_package['session_context' ]
available = context['available_tokens' ]
if available < 30000 :
trigger_context_compression()
if not can_agents_cooperate(
handoff_package['from_agent' ],
handoff_package['to_agent' ]
):
raise AgentCompatibilityError("Agents cannot cooperate" )
return True
Session Recovery (Enterprise v4.0.0)
Recovery Checkpoints
Task phase boundaries (before RED, GREEN, REFACTOR)
Agent handoffs
User interruptions
Token budget thresholds
Error conditions
Session timeouts
{
"checkpoint_id" : "ckpt_uuid" ,
"timestamp" : "2025-11-12T10:30:00Z" ,
"phase" : "GREEN" ,
"token_usage" : {
"used" : 85000 ,
"available" : 115000
} ,
"context_snapshot" : "...compressed snapshot..." ,
"session_id" : "sess_uuid" ,
"recovery_tokens_reserved" : 55000
}
Recovery Process (Enterprise v4.0.0)
State Restoration - Reload last valid checkpointContext Validation - Verify token budget sufficientSession Resumption - Use Agent SDK resume featureProgress Assessment - Determine what was completedContinuation Planning - Decide next steps with updated token budgetUser Notification - Inform user of recovery status
Memory State Synchronization
Memory Files (v4.0.0)
.moai/sessions/session-state.json - Current session metadata.moai/sessions/context-cache.json - Cached context for performance.moai/sessions/checkpoints/ - Saved recovery checkpoints.moai/sessions/token-usage.log - Token budget historyactive-tasks.md - TodoWrite task tracking Synchronization Protocol :
def sync_memory_files (session_state ):
"""Ensure memory files reflect current session state"""
update_session_metadata({
'session_id' : session_state.id ,
'token_usage' : session_state.token_budget,
'context_position' : session_state.context_position,
'last_sync' : timestamp()
})
sync_todowrite_tasks(session_state.active_tasks)
update_context_cache(compress_context(session_state.context))
log_token_usage({
'timestamp' : timestamp(),
'used' : session_state.tokens_used,
'available' : session_state.tokens_available,
'percent' : session_state.usage_percent
})
archive_old_checkpoints()
Best Practices (Enterprise v4.0.0)
Context Management
Use context-aware token budget tracking (Sonnet/Haiku 4.5 feature)
Create checkpoints before major operations
Apply progressive summarization for long workflows
Enable session persistence for recovery
Monitor token usage and plan accordingly
Use session forking for parallel exploration
Accumulate unlimited context history
Ignore token budget warnings
Skip state validation on recovery
Lose session IDs without saving
Mix multiple sessions without clear boundaries
Assume session continuity without checkpoint
Token Budget Optimization
Start new session per distinct task (fresh 200K tokens)
Use /context to identify expensive MCP servers
Compress context before handoffs
Keep reserve buffer (25-30% of tokens)
Monitor usage percent, not absolute tokens
Enable auto-summarization at 75% threshold
Let single conversation exceed 150K tokens
Keep all MCP servers enabled if not needed
Reuse sessions across fundamentally different tasks
Ignore available_tokens feedback
Store uncompressed context in memory files
Configuration Location : .moai/config/config.json
{
"session" : {
"persistence_enabled" : true ,
"auto_checkpoint" : true ,
"checkpoint_interval_minutes" : 10 ,
"recovery_strategy" : "progressive_summarization" ,
"context_budget" : {
"warning_threshold_percent" : 75 ,
"emergency_threshold_percent" : 85 ,
"reserve_tokens" : 55000
} ,
"forking_enabled" : true ,
"max_parallel_sessions" : 3
}
}
Debugging & Troubleshooting
Inspection Tools
/alfred:debug --show-session-state
/alfred:debug --show-token-budget
/alfred:debug --list-checkpoints
/alfred:debug --check-memory-sync
/alfred:debug --show-token-usage-log
Common Issues Issue Symptoms Solution Lost session Cannot resume conversation Check .moai/sessions/ for session IDs Token budget exceeded Model stops responding Use /context to identify heavy consumers, create new session Handoff failed Agent has wrong context Verify handoff package completeness before transfer Recovery stuck Cannot continue after interruption Restore from earlier checkpoint or start new session Memory drift Inconsistent information Run sync_memory_files() or check cache integrity
Version History Version Date Key Changes 4.0.0 2025-11-12 Enterprise context awareness, token budget tracking, Session SDK integration, Git 2.47+ support 1.1.0 2025-11-05 Session state management foundation 1.0.0 2025-10-01 Initial release
Related Skills
moai-alfred-context-budget - Token optimization deep divemoai-alfred-agent-guide - Multi-agent coordinationmoai-foundation-trust - State validation principlesmoai-foundation-git - Git session state tracking
Learn more in reference.md for detailed implementation guides, recovery procedures, advanced coordination patterns, and November 2025 API examples.
Skill Status : Production Ready | Last Updated: 2025-11-12 | Model Support: Sonnet 4.5, Haiku 4.5 | Enterprise v4.0.0
