memory-insights

You can analyze Claude Code session data for friction and satisfaction trends by reading per-session facets data to generate dashboards that surface outcome distributions and top friction types. This read-only view helps identify patterns for quicker improvement cycles.

Claude Code facets data is per-session analytics stored in the ~/.claude/usage-data/facets/ directory. You need to analyze it when you want to identify friction patterns, success rates, and satisfaction signals across multiple sessions to prioritize product improvements.

Yes, you can generate dashboards from Claude Code usage data without manual processing by requesting a friction trend dashboard. The tool reads the facets directory directly and gracefully handles its absence, returning a structured summary of available insights.

Yes, the session analytics tool works if the ~/.claude/usage-data/facets/ directory is missing. It gracefully handles the absence of the directory and returns a structured summary indicating what insights are currently available.

To identify which tasks cause the most friction across recent Claude Code sessions, generate a quick dashboard describing outcome distributions and top friction types. This correlates friction signals with goals to help product teams prioritize improvements.

The best way to track satisfaction trends in Claude Code sessions for product research is using a read-only, on-demand analysis of per-session facets data. This surfaces satisfaction signals and success correlations across multiple sessions for researchers and engineers.