What problem does it solve?
Hotplex-diagnostics helps you diagnose why the HotPlex Gateway appears healthy while users see no updates, slow responses, stuck tasks, or stalled streaming—by narrowing down failures from process state to the feedback pipeline and platform adapters.
Core Features & Use Cases
- Process & Port Verification: Check Gateway/WebChat/Admin worker subprocesses, PID-file consistency, and expected localhost listening to confirm the system is truly alive.
- Session State Consistency: Analyze HotPlex session counts and running/idle states across SQLite and on-disk Claude Code session files to detect orphaned/zombie/stale-resume conditions.
- Feedback Continuity & Stall Classification: Cross-validate Worker output timestamps against platform delivery timestamps to detect pipeline stalls, backpressure drops, adapter failures, or client disconnects (including silent stalls).
- Targeted Log Forensics: Search for meaningful WARN/ERROR patterns, reconstruct per-session timelines, and map symptoms to specific bridge/hub/platform/adapter failure modes.
- Adapter Connectivity Checks: Investigate Feishu/Slack reconnection patterns and potential connection leaks that can cause recurring streaming failures.
- Cross-platform Guidance: Provide Linux/macOS vs Windows command alternatives for practical debugging.
Quick Start
Ask to run hotplex-diagnostics when a user reports that HotPlex streaming is stuck or no longer produces output during a session, or before/after deployment to verify Gateway/Worker/adapters are functioning.