You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
{{ message }}
Repository navigation
Add efficiency metrics to classify_sessions() output #79
classify_sessions() categorizes sessions (debugging, development, research, maintenance, mixed) but doesn't include efficiency metrics. This makes it hard to answer "which session types burn the most context?"
Proposed Solution
Add efficiency fields to each session's classification output:
{
"session_id": "abc123",
"category": "debugging",
"compaction_count": 12,
"total_bytes": 5200000,
"files_read_multiple_times": 8,
"burn_rate": "high"# optional: simple high/medium/low based on compactions/hour
}
Value
Correlates session type with context efficiency, helping identify which kinds of work need better tooling or workflow changes.
Implementation
Join with efficiency data from get_session_efficiency() query
Activity
feat: Add limit parameters and efficiency metrics (closes #77, #78, #79)
feat: Add limit parameters and efficiency metrics (closes #77, #78, #79…
feat: Add compaction aggregation and pre-compaction pattern analysis (c…
evansenter commentedon Jan 10, 2026
Implemented in PR #80 - efficiency metrics added to classify_sessions()