- Comprehensive API documentation
- Migration examples (before/after)
- Specialized manager patterns
- Next steps for agent refactoring
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
- Create reusable SessionManager<T> with CRUD operations
- Support for session expiry, message limits
- Specialized managers for Domain and Server agents
- Reduce code duplication across 4 agents
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
- Remove unused config in commands.ts
- Remove deprecated formatDepositResult function in deposit-tool.ts
- Remove unused type imports in deposit-tool.ts
- Fix import path in deposit-agent.test.ts (src/deposit-agent → src/agents/deposit-agent)
This is a cleanup commit following the agent pattern unification work.
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
- Change field names to snake_case (user_id, collected_info, last_recommendation, created_at, updated_at, expires_at)
- Extract system prompts to constants (SERVER_EXPERT_PROMPT, SERVER_REVIEW_PROMPT)
- Add __PASSTHROUGH__/__SESSION_END__ marker support
- Change handler signature to match other agents (db, userId, userMessage, env, options)
- Add helper functions for consistency (createServerSession, isSessionExpired, addMessageToSession, hasServerSession)
- Update saveSe rverSession signature to not need userId separately
- Rename tools constant from serverExpertTools to serverExpertTools (camelCase)
- Change AI call parameter order for consistency
- Add performance logging
- Update openai-service.ts routing to use hasServerSession
- Update server-tool.ts to use new session creation helpers
- Update message-handler.ts and api/chat.ts field references
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
- Create troubleshoot_sessions table in D1
- Replace KV session storage with D1
- Unify field names to snake_case (user_id, collected_info, created_at, updated_at, expires_at)
- Add __PASSTHROUGH__/__SESSION_END__ marker support
- Change handler signature to match domain/deposit pattern
- Extract system prompt to constant TROUBLESHOOT_EXPERT_PROMPT
- Add hasTroubleshootSession() for routing
- Update openai-service.ts to use new D1-based functions
- Update troubleshoot-tool.ts to use D1 instead of KV
- Add TroubleshootSessionStatus type
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
- Add deposit session routing to openai-service.ts
- Convert deposit-tool to agent trigger (delegate to deposit-agent)
- Update CLAUDE.md with new agent architecture
Changes:
- openai-service.ts: Import and check hasDepositSession, route to processDepositConsultation
- deposit-tool.ts: Convert action → natural language → delegate to Deposit Agent
- CLAUDE.md: Update Core Services, Function Calling Tools, Data Layer, Deposit System sections
All tests passing (192/195), dev server starts successfully.
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
- Modify executeManageDomain to delegate to domain agent
- Add buildDomainMessage helper for action-to-message conversion
- Keep internal functions for agent tool execution
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
- Import domain agent functions
- Add domain session check before main AI processing
- Route to domain agent when session exists
- Handle __PASSTHROUGH__ for non-domain messages
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
- Add full conversation flow with session management
- Handle tool call execution
- Support __PASSTHROUGH__ and __SESSION_END__ markers
- Add hasDomainSession helper for routing
- Export executeDomainAction from domain-tool.ts
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
- Add executeDomainToolCall function
- Integrate with existing domain-tool.ts functions
- Map AI tool calls to domain actions
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
- Add domain expert system prompt
- Add domain tools for function calling
- Add callDomainExpertAI function
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
- Add type definitions for domain and deposit agent sessions
- Follow existing ServerSession pattern
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
- Create src/agents/ directory for agent modules
- Move server-agent.ts to new location
- Update import paths in all dependent files:
- openai-service.ts
- tools/server-tool.ts
- routes/handlers/message-handler.ts
- routes/api/chat.ts
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
Server Management:
- Fix /server command API auth (query param instead of header)
- Show server specs (vCPU/RAM/Bandwidth) in /server list
- Prevent AI from refusing server deletion based on expiration date
- Add explicit instructions in tool description and system prompt
Refund Display:
- Show before/after balance in server deletion refund message
- Format: 환불 전 잔액 → 환불 금액 → 환불 후 잔액
Other Changes:
- Add stopped status migration for server orders
- Clean up callback handler (remove deprecated code)
- Update constants and pattern utilities
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
- Add security.test.ts: 36 tests for webhook validation, rate limiting
- Add circuit-breaker.test.ts: 31 tests for state transitions
- Add retry.test.ts: 25 tests for exponential backoff
- Add api-helper.test.ts: 25 tests for API abstraction
- Add optimistic-lock.test.ts: 11 tests for concurrency control
- Add summary-service.test.ts: 29 tests for profile system
Total: 157 new test cases (222 passing overall)
- Fix setup.ts D1 schema initialization for Miniflare
- Update vitest.config.ts to exclude demo files
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
- Create Hono router in api.ts
- Convert 6 API endpoints to Hono format:
- GET /api/deposit/balance
- POST /api/deposit/deduct
- POST /api/test
- POST /api/chat
- POST /api/contact
- GET /api/metrics
- Use Hono CORS middleware for /contact
- Remove manual handleApiRequest and handleContactPreflight
- Integrate with main app via app.route('/api', apiRouter)
Benefits:
- Cleaner declarative routing (44 insertions, 48 deletions)
- Built-in CORS middleware
- Better code organization
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
- Replace stub implementations with actual API calls
- POST /api/provision/orders/{order_id}/start
- POST /api/provision/orders/{order_id}/stop
- Add proper validation, logging, and error handling
- Follow existing code patterns (callProvisionAPI, __DIRECT__)
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
P2-1: Tool selection fallback optimization
- Return only utility tools when no patterns match
- Reduces token usage by ~80% in fallback cases
P2-2: Minimum deposit amount validation
- Add MIN_DEPOSIT_AMOUNT = 1,000원
- Prevents spam with tiny deposits
P2-3: Standardize logging
- Replace console.log/error with structured logger
- bank-sms-parser.ts and security.ts
P2-4: Nameserver format validation
- Add validateNameservers() function
- Check minimum 2 NS, valid hostname format
- Clear error messages in Korean
P2-5: Optimistic lock error context
- Return specific error for version conflicts
- User-friendly message: "동시 요청으로 처리가 지연됨"
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
P1-1: Callback query error handling
- Add try-catch around domain registration and server order
- Send user-friendly error messages on failure
- Use answerCallbackQuery to acknowledge button clicks
- Add structured logging with createLogger
P1-2: Queue DLQ monitoring
- Add admin notification when server provisioning fails
- Update order status to 'failed' in database
- Include detailed context in notifications
- Apply rate limiting (1 notification per hour)
P1-3: Email handler error recovery
- Add admin notification when SMS parsing fails
- Include email preview in notifications
- Mask email addresses for privacy
- Add structured logging with emailLogger
Also add 'failed' status to ServerOrder type.
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
- Enhance OpenAI message types with tool_calls support
- Improve security validation and rate limiting
- Update utility tools and weather tool
- Minor fixes in deposit-agent and domain-register
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
- Add server-provision.ts for async server creation
- Add SERVER_PROVISION_QUEUE with DLQ for reliability
- Add cron job for auto-cleanup of pending orders (5min)
- Add server delete confirmation with inline keyboard
- Update types for server orders, images, and provisioning
- Add server tables to schema (server_orders, server_instances)
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
- Fix USD price display: all prices now show in KRW (₩)
- Add Korea region auto-detection: extracts region preference from user messages
- Fix low-spec recommendation for high-performance requirements:
- Add extractTechStack() to detect PostgreSQL, Redis, MongoDB keywords
- Enhance inferExpectedUsers() to consider tech stack complexity
- SaaS/B2B services now recommend 4GB+ RAM servers
- Fix __DIRECT__ tag appearing in output:
- Reorder message concatenation in server-agent.ts
- Add stripping logic in conversation-service.ts and api.ts
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
When user requests server recommendation while in an existing session
(e.g., selecting state), reset the session and start fresh instead of
continuing the old session context.
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
- Add response_format: { type: 'json_object' } for review mode to force JSON response
- Convert expectedDau and expectedConcurrent from string to number before API call
- Add enhanced KV session debugging with key names in logs
Fixes:
- AI returning plain text instead of JSON in review mode
- 400 error from recommend API due to string values in expected_users
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
Previously cancellation only worked in 'selecting' or 'ordering' states.
Now users can cancel server consultation at any stage using keywords:
취소, 다시, 처음, 리셋, 초기화, 다시 시작, 처음부터
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
- Add expectedDau and expectedConcurrent fields to ServerSession
- Update system prompts to explain DAU vs concurrent users concept
- AI now asks for clarification when users mention visitor counts
- Use concurrent users (5-10% of DAU) for server recommendations
- Update inference rules: personal=10, business=50 concurrent users
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
- Server Expert AI now reviews recommendation results before showing to user
- Changed flow: get recommendations first → AI reviews → show with comments
- AI provides specific advice based on actual recommended specs
- Reviews include: spec adequacy, bandwidth warnings, CDN suggestions
Before: AI gave generic advice without seeing recommendations
After: AI reviews actual results and gives contextual feedback
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
- Add cancel logic for selecting/ordering states
- Keywords: 취소, 다시, 처음
- Delete session and return to normal conversation
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
- Add 'selecting' and 'ordering' status to ServerSession
- Add lastRecommendation field to store recommendation results
- Keep session alive after recommendation (don't delete immediately)
- Add selection pattern matching (1번, 첫번째, 1번 선택 등)
- Add order confirmation message with inline buttons
- Add server_order/server_cancel callback handlers
- Add ServerOrderKeyboardData type for button data
Flow: recommend → select number → confirm with buttons → order/cancel
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
