Logs Explorer PRO
The Logs Explorer gives you full visibility into your OpenClaw instance. Stream logs in real time directly from your VPS via SSH, search with keyword highlighting, filter by severity level, detect recurring patterns automatically, and configure alerting rules that notify you when things go wrong.
Real-Time Log Streaming
Logs are streamed live from your VPS over a secure SSH connection. As your agents process messages and handle requests, each log line appears in the explorer within seconds. The stream stays open as long as you have the page active, so you can monitor behavior during deployments, debugging sessions, or high-traffic periods.
Auto-refresh intervals are configurable at 5s, 10s, 30s, or 60s. Choose a faster interval for active debugging or a slower one for passive monitoring. You can also pause the stream entirely to inspect a specific moment in the log output without new entries pushing it out of view.
Search and Keyword Highlighting
Type any search term into the search bar to filter the visible log lines instantly. Matching keywords are highlighted in the output so you can spot them at a glance. Search works across all log fields including timestamps, log levels, agent names, and message content.
Combine search with level filters to narrow results further. For example, search for "timeout" while filtering to error level to find only timeout-related errors without wading through informational noise.
Level Filtering
Filter logs by severity level to focus on what matters. The four supported levels are:
- Error — Failures that need immediate attention (failed requests, crashes, unhandled exceptions)
- Warn — Potential issues that may degrade performance or indicate misconfiguration
- Info — Standard operational messages (agent started, channel connected, message processed)
- Debug — Verbose output useful during development and troubleshooting
Multiple levels can be selected simultaneously. The level filter persists across page refreshes and works in combination with the search bar.
Saved Searches and Views
Save frequently used filter combinations as named views for instant access. Each saved view stores the search query, selected log levels, line count, and date/time range. You can create up to 20 saved views per account.
Set any saved view as your default view, and it will load automatically every time you open the Logs Explorer. This is useful for teams that always want to start with a specific filter, such as "errors only in the last hour."
Log Pattern Detection
The pattern detection engine automatically scans your log stream and identifies recurring patterns. When the same error or warning appears repeatedly, the system groups those occurrences and surfaces them as a detected pattern with a count, first/last occurrence timestamps, and a representative sample.
Pattern detection helps you identify systemic issues that might be missed when reading logs line by line. A single timeout error is easy to overlook; fifty of them in an hour is a pattern that demands investigation.
Log Alerting
Configure alerting rules to get notified when specific conditions are met. Each rule consists of three parts:
- Condition — What to match (keyword, log level, or pattern)
- Threshold — How many occurrences within a time window trigger the alert
- Notification — Where to send the alert (email, webhook, or dashboard notification)
For example, you can create a rule that sends a webhook notification if the word "OutOfMemory" appears more than 5 times in a 10-minute window. Alert rules are evaluated continuously against the incoming log stream.
Date/Time Filtering
Narrow your log view to a specific time range using the date/time picker. Select a start and end timestamp to view only the logs generated during that window. This is essential for post-incident analysis when you need to reconstruct what happened during a specific period.
Line Count and Download
Control how many log lines are displayed using the line count selector. Choose from predefined values or enter a custom count. For offline analysis or sharing with your team, download the currently filtered log output as a .txt file. The download respects all active filters, so you get exactly what you see on screen.
Log Forwarding
Forward your logs to external logging services for long-term storage, cross-service correlation, or integration with your existing observability stack. Configure a forwarding destination and ClawHQ will continuously ship log data to your external service alongside the in-dashboard experience.
Log forwarding works independently of the Logs Explorer. Even if you never open the explorer, forwarded logs are delivered to your configured destination.
Related Documentation
- Pro Features Overview — Full list of Pro capabilities
- Analytics — Aggregate insights from your conversation data
- Audit Log — Track administrative actions with tamper-proof records
- Webhooks — Send log alerts to external services via webhook