ctx-monitor

Run modular audits on execution traces

Manage ctx-monitor per-project configuration

Launch interactive web dashboard with real-time metrics

Compare traces between sessions to identify regressions

Diagnose and fix ctx-monitor installation problems

Fast trace validation agent using Haiku for cost-efficient basic checks. Use this agent for quick schema validation, event counting, and integrity checks. IMPORTANT: This agent ONLY performs deterministic validation tasks. If complex analysis is needed, it will stop and suggest using trace-analyzer. <example> Context: User wants to quickly check if traces are valid user: "validate my traces" assistant: "I'll use the quick-validator agent to perform fast validation of your trace files." <commentary> Quick validation request triggers the quick-validator agent for schema and integrity checks. </commentary> </example> <example> Context: User wants to count events before detailed analysis user: "how many events in my traces?" assistant: "I'll use the quick-validator agent to count and categorize your trace events." <commentary> Event counting is a deterministic task perfect for the quick-validator with Haiku. </commentary> </example> <example> Context: User suspects corrupted trace files user: "check if my trace files are corrupted" assistant: "I'll use the quick-validator agent to check file integrity and JSON structure." <commentary> Corruption detection is a structural check that doesn't require complex reasoning. </commentary> </example>

Use this agent when the user needs to analyze ctx-monitor execution traces and identify issues. This agent triggers after `/ctx-monitor:report` runs or when the user requests trace analysis. <example> Context: User just ran /ctx-monitor:report and sees errors in the output user: "analyze traces" assistant: "I'll use the trace-analyzer agent to perform deep analysis of the execution traces." <commentary> The user wants to understand what went wrong in their traces. The trace-analyzer agent will read the trace files, identify patterns, and provide actionable recommendations. </commentary> </example> <example> Context: User is debugging intermittent hook failures user: "find issues in execution" assistant: "I'll use the trace-analyzer agent to investigate the execution traces for anomalies and intermittent failures." <commentary> The phrase "find issues in execution" indicates the user wants trace analysis. The agent will search for patterns like hooks that sometimes fire and sometimes don't. </commentary> </example> <example> Context: User ran ctx-monitor:report and noticed high error rates user: "why are there so many errors in my traces?" assistant: "I'll use the trace-analyzer agent to investigate the error patterns and identify root causes." <commentary> The user is asking about trace errors. The agent will analyze error patterns, correlate them with tool calls, and suggest fixes. </commentary> </example> <example> Context: User suspects something is wrong with their ctx-monitor setup user: "debug ctx-monitor logs" assistant: "I'll use the trace-analyzer agent to examine the ctx-monitor logs and identify any issues." <commentary> Direct request to debug ctx-monitor logs triggers the trace-analyzer agent for comprehensive log analysis. </commentary> </example>

This skill should be used when the user asks about "interpreting ctx-monitor traces", "understanding execution logs", "reading trace files", "what do ctx-monitor events mean", "debugging trace output", or needs guidance on trace file formats, event types, common failure patterns, or troubleshooting ctx-monitor output.

9 hooks across 9 events