Generate insights summary report from conversation analysis
From codesurf-insightsnpx claudepluginhub jasonkneen/codesurf-insights[week|month|all]/reportGenerates reports on task execution, progress, and metrics including executive summaries, sprint burndowns, standups, performance analysis, and dependencies.
/reportGenerates reports on task execution, progress, and metrics including executive summaries, sprint burndowns, standups, performance analysis, and dependencies.
/reportGenerates formatted terminal report of AI experiments with stats, top tools, rating distribution, tags, recent activity, and latest details. Supports optional tool, tag, days, rating filters.
/reportGenerates performance report from benchmark results with stats, text histograms, endpoint rankings, SLA checks, throughput, error analysis, regressions, recommendations, and executive summary as markdown.
/reportGenerates a project health report with git activity, code quality, dependency, test, and documentation metrics including score and top issues.
/reportGenerates SEO and GEO performance report for a domain over a time period, aggregating metrics across channels with trends, analysis, and prioritized recommendations. Supports comparison periods and executive format.
Create a comprehensive report summarizing insights extracted from conversation analysis. Outputs both to terminal and saves to file.
week (default): Report on last 7 days of analysismonth: Report on last 30 daysall: Report on all available analysis dataYYYY-MM-DD: Report from specific dateGather Analysis Data Read from:
~/.claude/insights/reports/~/.claude/insights/archive/*/Aggregate by Dimension Group insights by the 6 analysis dimensions:
Calculate Statistics Compute metrics:
Identify Patterns Find recurring themes:
Output to Terminal Display formatted summary:
CodeSurf Insights Report
════════════════════════════════════════════
Period: 2024-01-08 to 2024-01-15
STATISTICS
──────────────────────────────────
Conversations: 47
Messages: 2,891
Detritus archived: 634 (22%)
Insights extracted: 89
TOP LESSONS LEARNED
──────────────────────────────────
1. Always verify file exists before editing (seen 7x)
2. Use Task agent for multi-file searches (seen 5x)
3. Check imports before adding new ones (seen 4x)
EFFICIENCY OPPORTUNITIES
──────────────────────────────────
1. Could save ~3 turns by reading full file first
2. Parallel tool calls underutilized
3. Over-explaining in responses
COMMON ASSUMPTIONS (Incorrect)
──────────────────────────────────
1. Assuming TypeScript when user has JavaScript
2. Assuming npm when user prefers pnpm
3. Assuming main branch instead of checking
LANGUAGE ADDITIONS
──────────────────────────────────
- "PR" → Pull Request
- "WIP" → Work In Progress
- "LGTM" → Looks Good To Me
TREND
──────────────────────────────────
↑ Efficiency improved 12% from last period
↓ More assumptions this week (investigate)
→ Detritus ratio stable at ~22%
Save Report File
Write detailed version to:
~/.claude/insights/reports/YYYY-MM-DD-HHmm.md
Include:
# CodeSurf Insights Report
Generated: 2024-01-15 14:30
Period: 2024-01-08 to 2024-01-15
## Summary
- Conversations: 47
- Messages: 2,891
- Insights: 89
## Lessons Learned
### 1. Verify file exists before editing
**Frequency**: 7 occurrences
**Example**: In conversation abc123, attempted to edit...
**Recommendation**: Add file existence check to editing workflow
### 2. Use Task agent for parallel searches
...
all quarterly for comprehensive review