All skills
bitwarden avatar

/analyzing-git-sessions

@2be0603 official
by bitwardenbitwarden/ai-plugins155 stars
20

Analyzes git commits and changes within a timeframe or commit range, providing structured summaries for code review, retrospectives, work logs, or session documentation.

Use this Skill: https://skilld.dev/gh/bitwarden/ai-plugins/analyzing-git-sessions

This session only. Nothing lands on disk.

README.md

≈2.5k tokens on demand. Your agent reads this file only when SKILL.md points to it.

Analyzing Git Sessions

Analyzes git commits and changes within a timeframe or commit range, providing structured summaries for various use cases.

What It Does

This skill analyzes git repository activity and produces structured summaries including:

  • Commit history with messages and authors
  • File change statistics (additions, deletions, modifications)
  • Identification of key files by change magnitude
  • Optional code diffs for review
  • Multiple output formats for different use cases

When to Use

Use this skill when you need to:

  • Review your work: "What did I accomplish in the last 2 hours?"
  • Prepare code reviews: Generate PR descriptions and identify review focus areas
  • Write standups: "What did I work on yesterday?"
  • Document sessions: Capture work done during a coding session
  • Analyze branches: "What changes are in my feature branch?"
  • Track activity: Weekly or monthly contribution summaries

How to Use

Natural Language Invocations

Time-based analysis:

  • "Analyze my git changes from the last 2 hours"
  • "Show me what I committed today"
  • "Summarize my work since 10am"
  • "What changed in the last hour?"

Commit range analysis:

  • "Show me changes between abc123 and def456"
  • "Analyze commits from HEAD~5 to HEAD"
  • "Compare feature-branch to main"

With specific focus:

  • "Show me authentication module changes from today"
  • "What test files did I modify this week?"
  • "Analyze changes in the app directory since yesterday"

Output format requests:

  • "Generate a code review summary for my last 3 commits"
  • "Give me a concise summary of today's work"
  • "Show detailed diffs for my morning commits"

Direct Skill Invocation

/skill analyzing-git-sessions

The skill will prompt you for:

  1. Time range or commit range
  2. Optional path filters
  3. Desired output depth (concise/detailed/code review)

Output Formats

1. Concise Summary (Default)

Best for: Quick overview, standup reports, work logs

Includes:

  • Commit count and authors
  • Files changed summary (added/modified/deleted)
  • Net line changes
  • Top 5-10 files by change size
  • Commit messages

Example:

Git Session Summary
Range: 2 hours ago to now
Commits: 8
Files Changed: 12 modified, 2 added
Net Changes: +340 -120 lines

Top Files:
1. AuthViewModel.kt (+85 -30)
2. LoginScreen.kt (+60 -15)
...

2. Detailed Summary

Best for: Session documentation, deeper review

Includes everything from Concise, plus:

  • Full file list with statistics
  • Author breakdown
  • Diffs for top 3-5 files
  • File categorization (source, test, config)

3. Code Review Format

Best for: PR preparation, team reviews

Includes:

  • Suggested PR title (from commits)
  • Changes grouped by module/directory
  • Scope summary
  • Test coverage assessment
  • Formatted commits suitable for PR description
  • Key diffs organized by module

Configuration Options

Output Depth

Control level of detail:

  • Concise: Stats and commit list only (~500-1000 tokens)
  • Detailed: Stats + top file diffs (~2000-5000 tokens)
  • Code Review: Comprehensive with module grouping (~3000-8000 tokens)

The skill automatically suggests appropriate depth based on session size.

Path Filters

Narrow analysis to specific paths:

  • "Analyze changes in app/src/main/kotlin/auth/"
  • "Show only test file changes"
  • "Focus on gradle files"

Author Filters

For multi-contributor analysis:

  • "Show only my commits from today"
  • "What did the team change this week?"

Use Cases

1. Daily Standup Preparation

Request: "Summarize my git changes from yesterday"

Result: Concise list of commits and key files for standup discussion

2. Code Review Preparation

Request: "Generate code review summary for feature-auth-biometrics branch"

Result: PR-ready description with:

  • What changed and why
  • Files organized by module
  • Test coverage notes
  • Suggested review focus areas

3. Session Documentation

Request: "Detailed analysis of my work in the last 3 hours"

Result: Comprehensive summary for retrospective or work log:

  • What was accomplished
  • Which files were touched
  • Magnitude of changes
  • Key code diffs

4. Weekly Activity Summary

Request: "Show me all my commits from the last 7 days"

Result: Week-at-a-glance view of your contributions

5. Branch Comparison

Request: "Compare feature-branch to main"

Result: What's in the feature branch that's not in main

  • Commits to be merged
  • Files that will change
  • Potential conflicts

6. Retrospective Git Data

Request: (Called by retrospecting skill during analysis)

Result: Git metrics for retrospective:

  • Session scope (commit count, file count)
  • Change velocity
  • Areas of focus (which modules touched)

Tips for Best Results

Be Specific About Timeframe

Good:

  • "Last 2 hours"
  • "Since 2025-10-23 14:00"
  • "Today from 9am to 5pm"

Less specific (still works, but may need clarification):

  • "Recent changes" → skill will ask "how recent?"
  • "My work" → skill will ask "from when?"

Request Appropriate Depth

For quick checks: "Give me a concise summary" For thorough review: "Detailed analysis with diffs" For PR prep: "Code review format"

The skill will recommend depth based on session size, but you can override.

Use Filters for Large Sessions

If you worked on multiple areas:

  • "Show only authentication changes from today"
  • "Analyze ViewModel changes in the last week"

This keeps output focused and context-efficient.

Output Examples

Example 1: Quick Work Log

Input: "What did I work on this morning?"

Output:

Git Session Summary (9:00am - 12:00pm)

Commits: 6
Files: 9 modified, 1 added
Changes: +220 -65 lines

Commits:

- Add biometric authentication support
- Refactor LoginViewModel state handling
- Update authentication tests
- Fix credential manager integration
- Add error handling for biometric failures
- Update documentation

Key Files:

1. BiometricAuthManager.kt (+85 -15)
2. LoginViewModel.kt (+60 -25)
3. AuthenticationTests.kt (+45 -20)

Example 2: Code Review Ready

Input: "Generate PR summary for my feature branch"

Output:

Pull Request Summary

Title: Add Biometric Authentication Support

Overview:

- 12 commits over 3 days
- 15 files changed across 3 modules
- Adds biometric authentication flow with fallback

Commits:

1. Implement BiometricAuthManager for authentication
2. Add biometric prompt UI components
3. Integrate with CredentialManager API
   ...

Changes by Module:

app (8 files):

- BiometricAuthManager.kt: Core biometric logic
- LoginScreen.kt: UI integration
- AuthViewModel.kt: State management

core (4 files):

- CredentialProvider.kt: Credential management
- SecurityUtils.kt: Security helpers

tests (3 files):

- BiometricAuthTests.kt: Unit tests
- LoginFlowTests.kt: Integration tests

Test Coverage:
✓ 2 new test files added
✓ 45 new test cases
⚠ Consider: Error handling edge cases

[Diffs for key files would follow]

Integration with Other Skills

This skill integrates with several workflows and can be invoked standalone or as part of other processes:

Retrospecting Skill

  • Invoked during Step 2 (Gather Data - Git Analysis)
  • Provides commit history and change stats for retrospective analysis

Code Review Preparation

  • Generate PR descriptions
  • Identify review focus areas
  • Highlight test coverage changes

Daily Standup Reports

  • "What did I work on yesterday?"
  • Quick summary of activity

Work Log Generation

  • Weekly/monthly activity summaries
  • Contribution tracking

Performance Notes

Context Efficiency

  • Concise mode: ~500-1K tokens (very efficient)
  • Detailed mode: ~2-5K tokens (moderate)
  • Code Review mode: ~3-8K tokens (comprehensive)

For large sessions (>30 files), the skill automatically uses concise mode to prevent context overflow.

Speed

  • Git commands are fast (typically <1 second)
  • Analysis and formatting: 5-10 seconds
  • Total time: Usually <30 seconds

Limitations

Cannot:

  • Analyze uncommitted changes (use git status or git diff directly)
  • Access remote branches not fetched locally
  • Interpret semantic meaning of changes (shows what changed, not why)
  • Generate commits or modify git history

Works best with:

  • Clean commit history with meaningful messages
  • Commits pushed regularly (not one giant commit)
  • Reasonable session sizes (<100 files changed)

Troubleshooting

"No commits found in range"

  • Check your time syntax: "since 2 hours ago" not "2 hours"
  • Verify commits exist: git log --since="your timeframe"
  • Try absolute time: "since 2025-10-23 10:00"

"Too many files changed, showing summary only"

  • This is context protection (>30 files)
  • Request specific paths: "Show only app/ changes"
  • Use concise mode explicitly
  • Split analysis into multiple focused queries

"Commit range invalid"

  • Check commit hashes exist: git log abc123
  • Use correct range syntax: start..end (two dots)
  • Ensure branch names are correct

Output seems incomplete

  • Specify "detailed" mode explicitly
  • Check if session size triggered automatic concise mode
  • Request specific files: "Show diffs for MainActivity.kt"

Version History

v1.0 (2025-10-23):

  • Initial release
  • Support for time ranges and commit ranges
  • Three output formats (concise, detailed, code review)
  • Context-aware depth selection
  • Integration with retrospecting skill

Feedback

Found an issue or want a feature?

  • Update .claude/skills/analyzing-git-sessions/SKILL.md
  • Modify this README for documentation improvements
  • Suggest new output formats or filters

Maintained By: @team-ai-sme Related Skills: retrospecting, reviewing-changes

Source: SKILL.md on GitHub

No alerts14d4 checks · Risk SAFE
  • Gen Agent Trust Hub14d

    The skill facilitates git repository analysis through shell commands. While it includes input validation, it is vulnerable to indirect prompt injection from repository content and executes shell commands based on user-supplied strings.

  • Socket14d

    No alerts

  • Snyk14d

    Risk: LOW · No issues

  • Runlayer7mo

    2/3 files flagged

Signed by skilld at 2be0603. This ties the file your Agent reads to that commit on GitHub. It does not review the instructions.

Last checked against GitHub yesterday.

Activeupdated 10 months ago
  • Git/VCS
  • commits
  • code-review
  • retrospectives
  • work-logs
  • session-analysis
  • diffs
  • statistics

README badge

README badge for bitwarden/ai-plugins/analyzing-git-sessions

Analyzes git commits and changes within a specified timeframe or commit range, producing structured summaries with commit history, file statistics, and optional diffs. Useful for code reviews, retrospectives, work logs, and pull request descriptions where you need a quick overview of what changed and by how much.

Generated from the current SKILL.md.

What time formats does this skill accept?
Relative formats like 'last 2 hours' or 'since 10am', absolute timestamps like '2025-10-23 14:00', and commit ranges like 'abc123..def456' or 'HEAD~5..HEAD'.
Does this skill show diffs for all changed files?
No. It uses progressive disclosure: concise mode shows stats only, detailed mode shows diffs for top 3-5 files, and code review mode shows diffs grouped by module. Large sessions (>30 files, >2000 lines) default to concise mode to avoid context overflow.
Can I filter by specific authors or file paths?
Yes. The skill accepts optional filters for specific paths, authors, or file types as part of the input.
What does 'code review format' output include?
It provides an overview with inferred PR title, commits suitable for a PR description, changes grouped by module, key diffs, and test coverage assessment.

Generated from the current SKILL.md. These answers refresh after source changes.