CtrlK
BlogDocsLog inGet started
Tessl Logo

insights-report

Generate a comprehensive cross-session insights report by analyzing all captured session logs in `.github/sessions/`. Use when reviewing patterns across sessions or preparing retrospectives.

Install with Tessl CLI

npx tessl i github:0xrabbidfly/eric-cartman --skill insights-report
What are skills?

83

1.62x

Quality

76%

Does it follow best practices?

Impact

99%

1.62x

Average score across 3 eval scenarios

Optimize this skill with Tessl

npx tessl skill review --optimize ./.github/skills/insights-report/SKILL.md
SKILL.md
Review
Evals

Insights Report Skill

Generate a comprehensive insights report by analyzing all captured session logs.

Usage

/insights-report

Optional: Specify date range

/insights-report --from 2026-02-01 --to 2026-02-09

Instructions

You are analyzing all session logs in .github/sessions/ to generate a comprehensive insights report (project-specific).

Step 1: Discover Sessions

  1. Find all session files (PowerShell):

    Get-ChildItem -Path .github/sessions -Filter *.md -Recurse -File
  2. Count and categorize:

    • Total sessions
    • Date range (earliest to latest)
    • Sessions per day/week
  3. If no sessions found:

    • Tell user: "No session logs found in .github/sessions/. Run /session-log end to capture your first session!"
    • Stop here

Step 2: Parse Session Data

For each session file, extract:

  • Session name (from filename and/or title)
  • Date
  • Duration (if captured)
  • Primary goal (from "Primary Goal" section)
  • Outcome (from "Outcome" section: fully/mostly/unclear)
  • Files modified (from "Files Modified" section)
  • Commands run (from "Commands Executed" section)
  • Tool usage (from "Tool Usage" section)
  • Challenges (from "Challenges & Friction" section)
  • Insights (from "Insights & Learnings" section)
  • What worked (from "What Worked Well" section)
  • What didn't work (from "What Didn't Work" section)
  • Incremental logs (any challenge/insight/metric entries)

Step 3: Analyze Patterns

Aggregate across all sessions:

Project Areas

  • Group sessions by topic/goal similarity
  • Identify major workstreams (e.g., "TypeScript type fixes", "UI development", "Backend work")
  • Count sessions per area

Goals & Outcomes

  • Most common goals (top 5-10)
  • Outcome distribution (% fully/mostly/unclear achieved)
  • Sessions with unclear outcomes (incomplete work)

Tool Usage Patterns

  • Most-used tools across sessions
  • Heavy read patterns (investigation-heavy sessions)
  • Multi-file edit patterns (large refactoring sessions)
  • Terminal-heavy sessions (command-line work)

Friction Points

  • Most common challenges
  • Recurring issues
  • Patterns in what doesn't work
  • Misunderstandings or clarifications needed

What Works

  • Successful workflows
  • Efficient approaches
  • Patterns in fully-achieved sessions
  • Insights that led to success

File & Command Patterns

  • Most frequently modified files
  • Common command sequences
  • Test/typecheck/build patterns

Time Patterns

  • Average session duration
  • Long vs. short sessions
  • Time-of-day patterns (if timestamps available)

Step 4: Generate Insights Report

Create a comprehensive markdown report with these sections:

# Project Insights Report

**Generated:** 2026-02-09
**Sessions Analyzed:** 42
**Date Range:** 2026-01-15 to 2026-02-09
**Total Duration:** ~84 hours

---

## Executive Summary

<2-3 paragraphs summarizing the key findings>

**What's working:** <Brief summary>

**What's hindering:** <Brief summary>

**Top recommendation:** <One actionable suggestion>

---

## Project Areas

You've been working in these main areas:

### 1. TypeScript Type System Fixes (18 sessions)
Description of work done, patterns observed, outcomes

### 2. UI Development (12 sessions)
Description of work done, patterns observed, outcomes

### 3. Backend & API Work (8 sessions)
Description of work done, patterns observed, outcomes

<Additional areas as discovered>

---

## Top Goals

Your most common objectives across sessions:

1. **Fix TypeScript type errors** (18 occurrences)
   - Outcome: 12 fully achieved, 4 mostly achieved, 2 unclear
   - Average duration: ~2.5 hours

2. **Review and improve UI** (12 occurrences)
   - Outcome: 7 fully achieved, 3 mostly achieved, 2 unclear
   - Average duration: ~1.5 hours

3. **Debug authentication issues** (6 occurrences)
   - Outcome: 4 fully achieved, 2 mostly achieved
   - Average duration: ~3 hours

<Continue for top 5-10 goals>

---

## What's Working Well

### Impressive Workflows

#### 1. Systematic Type Error Fixing
<Description of what makes this work well>

**Example sessions:**
- 2026-02-05: fix-typescript-errors.md
- 2026-02-07: resolve-auth-types.md

#### 2. <Another successful pattern>
<Description>

### Successful Patterns
- <Pattern that appears in fully-achieved sessions>
- <Another effective approach>
- <Efficient workflow observed>

---

## Where Things Go Wrong

### Friction Categories

#### 1. Sessions Ending Before Verification (8 sessions)
<Description of the pattern>

**Examples:**
- 2026-02-03: typecheck-auth.md - Fixed 39 errors but no final verification
- 2026-02-06: ui-review.md - Investigation started but no resolution

**Suggestion:** <How to improve>

#### 2. <Another friction pattern>
<Description>

**Examples:**
<Specific sessions>

**Suggestion:** <How to improve>

### Common Challenges
- **TypeScript type errors** (mentioned in 15 sessions)
- **Missing dependencies** (mentioned in 8 sessions)
- **UI styling issues** (mentioned in 12 sessions)

---

## Tool Usage Patterns

- **Most-used tool:** Read (1,234 times) - Heavy investigation work
- **Multi-file edits:** Edit used 456 times across 89 files
- **Command-line work:** Terminal used 234 times

**Patterns:**
- Sessions with >20 Read calls tend to have unclear outcomes (investigation-heavy)
- Sessions with <10 Edit calls more likely to be fully achieved (focused work)
- Typecheck commands run in 85% of sessions

---

## File Patterns

### Most Modified Files
1. `src/auth/login.ts` (12 sessions)
2. `src/models/user.ts` (10 sessions)
3. `src/components/Dashboard.tsx` (9 sessions)
<Continue for top 10>

### Common Command Sequences
1. `npx tsc --noEmit` → fix files → `npx tsc --noEmit` (typecheck loop)
2. `git status` → file changes → `git add` → `git commit` (commit flow)
3. `npm test` → fix tests → `npm test` (test-driven fixes)

---

## Key Insights & Learnings

Aggregated from your session logs:

- "<Insight from session X>"
- "<Learning from session Y>"
- "<Pattern discovered in session Z>"

<List top 10-15 insights>

---

## Suggestions

### Quick Wins

#### 1. <Actionable suggestion based on friction>
**Why:** <Explanation based on data>
**How:** <Specific steps>
**Expected impact:** <What improves>

#### 2. <Another suggestion>
<Details>

### Process Improvements

#### 1. <Workflow improvement>
**Current pattern:** <What happens now>
**Suggested pattern:** <What should happen>
**Example:** <How to implement>

#### 2. <Another improvement>
<Details>

### .github/copilot-instructions.md Additions

Based on your patterns, add these to `.github/copilot-instructions.md`:

```markdown
## <Section suggestion 1>
<Recommended content based on session patterns>
## <Section suggestion 2>
<Recommended content>

Features to Try

1. <Copilot / VS Code feature recommendation>

Why for you: <Based on your usage patterns> How to use: <Specific example>

2. <Another feature>

<Details>

Session Outcomes

  • Fully Achieved: 28 sessions (67%)
  • ⚠️ Mostly Achieved: 10 sessions (24%)
  • Unclear: 4 sessions (9%)

Characteristics of fully-achieved sessions:

  • <Pattern 1>
  • <Pattern 2>
  • <Pattern 3>

Characteristics of unclear-outcome sessions:

  • <Pattern 1>
  • <Pattern 2>

Timeline

Sessions by Week

  • Week of 2026-01-15: 8 sessions
  • Week of 2026-01-22: 12 sessions
  • Week of 2026-01-29: 15 sessions
  • Week of 2026-02-05: 7 sessions

Most Active Days

  1. 2026-01-28: 5 sessions
  2. 2026-02-02: 4 sessions
  3. 2026-01-23: 4 sessions

Appendix

All Sessions

<Chronological list of all session files analyzed>

Generated by Copilot Chat /insights-report skill Analyzed {N} sessions from .github/sessions/

Repository
0xrabbidfly/eric-cartman
Last updated
Created

Is this your skill?

If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.