2 Commits

Author	SHA1	Message	Date
vikingowl	cdf95002fc	feat(phase9): implement WebSocket transport and failover system ... Implements Phase 9: Remoting / Cloud Hybrid Deployment with complete WebSocket transport support and comprehensive failover mechanisms. **WebSocket Transport (remote_client.rs):** - Added WebSocket support to RemoteMcpClient using tokio-tungstenite - Full bidirectional JSON-RPC communication over WebSocket - Connection establishment with error handling - Text/binary message support with proper encoding - Connection closure detection and error reporting **Failover & Redundancy (failover.rs - 323 lines):** - ServerHealth tracking: Healthy, Degraded, Down states - ServerEntry with priority-based selection (lower = higher priority) - FailoverMcpClient implementing McpClient trait - Automatic retry with exponential backoff - Circuit breaker pattern (5 consecutive failures triggers Down state) - Background health checking with configurable intervals - Graceful failover through server priority list **Configuration:** - FailoverConfig with tunable parameters: - max_retries: 3 (default) - base_retry_delay: 100ms with exponential backoff - health_check_interval: 30s - circuit_breaker_threshold: 5 failures **Testing (phase9_remoting.rs - 9 tests, all passing):** - Priority-based server selection - Automatic failover to backup servers - Retry mechanism with exponential backoff - Health status tracking and transitions - Background health checking - Circuit breaker behavior - Error handling for edge cases **Dependencies:** - tokio-tungstenite 0.21 - tungstenite 0.21 All tests pass successfully. Phase 9 specification fully implemented. 🤖 Generated with [Claude Code](https://claude.com/claude-code) Co-Authored-By: Claude <noreply@anthropic.com>	2025-10-10 20:43:21 +02:00
vikingowl	33d11ae223	fix(agent): improve ReAct parser and tool schemas for better LLM compatibility ... - Fix ACTION_INPUT regex to properly capture multiline JSON responses - Changed from stopping at first newline to capturing all remaining text - Resolves parsing errors when LLM generates formatted JSON with line breaks - Enhance tool schemas with detailed descriptions and parameter specifications - Add comprehensive Message schema for generate_text tool - Clarify distinction between resources/get (file read) and resources/list (directory listing) - Include clear usage guidance in tool descriptions - Set default model to llama3.2:latest instead of invalid "ollama" - Add parse error debugging to help troubleshoot LLM response issues The agent infrastructure now correctly handles multiline tool arguments and provides better guidance to LLMs through improved tool schemas. Remaining errors are due to LLM quality (model making poor tool choices or generating malformed responses), not infrastructure bugs. 🤖 Generated with [Claude Code](https://claude.com/claude-code) Co-Authored-By: Claude <noreply@anthropic.com>	2025-10-09 19:43:07 +02:00