Commit Tracker Agent

Purpose

The Commit Tracker agent monitors git commit history, intelligently maps commits to Jira issues, and posts comprehensive commit summaries to Jira with links to related Confluence documentation. This agent creates a complete development audit trail and ensures all work is properly tracked.

Core Responsibilities

Commit Discovery: Identify relevant commits from git history
Issue Mapping: Map commits to Jira issues using multiple strategies
Detail Extraction: Extract comprehensive commit metadata
Summary Generation: Create detailed commit reports
Jira Integration: Post commit summaries as Jira comments
Confluence Linking: Connect commits to related documentation

Commit-to-Issue Mapping Algorithm

Multi-Strategy Mapping

Direct Key Match (1.0): Issue key in commit message (PROJ-123)
File Path Match (0.7): Changed files match issue scope
Temporal Proximity (0.5): Commit timestamp near issue activity
Semantic Analysis (0.4): Commit message similar to issue summary

File patterns from issue description, custom fields, or labels. Score: 0.0-0.7 based on match %

Temporal: Score based on commit time vs issue activity (< 1hr: 0.5, < 4hr: 0.4, < 24hr: 0.3, < 72hr: 0.2, else: 0.1)

Semantic: Extract keywords from commit message and issue summary/description, calculate overlap (0.0-0.4)

Final Scoring: Direct match → immediate return. Otherwise combine file + temporal + semantic scores. Filter by min confidence (0.6)

Commit Detail Extraction

Fields: sha, short_sha, author, email, date, timestamp, subject, body, parent, files, stats, branches, is_merge

File Info: path, status (added/modified/deleted), additions, deletions, binary flag

Jira Comment Template

Format: Author, date, branch, commit message, files changed table (+/-), GitHub link, Confluence links, complexity/risk

Variables: short_sha, author_name/email, commit_date, branch_name, subject, body, additions, deletions, file_table_rows

Batch Processing

Process: Group commits by issue → Map each → Sort by timestamp → Generate batch comment → Post to Jira

Batch Comment: Period, total commits, authors, individual summaries, files changed, lines added/deleted

Integration with Smart Commit Validator

Workflow Integration

commit_tracking_workflow:

  1_commit_made:
    trigger: git-commit
    action: validate_commit
    agent: smart-commit-validator

  2_validation_passed:
    trigger: validation-success
    action: track_commit
    agent: commit-tracker

  3_map_to_issues:
    action: map_commit_to_issues
    strategies:
      - direct_key_match
      - file_path_match
      - temporal_proximity
      - semantic_analysis

  4_extract_details:
    action: extract_commit_metadata
    fields:
      - files_changed
      - author_info
      - timestamps
      - parent_commits
      - branch_info

  5_generate_comment:
    action: generate_jira_comment
    template: standard_commit_summary
    include:
      - commit_details
      - file_changes
      - github_links
      - confluence_refs

  6_post_to_jira:
    action: post_comment
    target: jira_issue
    on_success: update_tracking_db

  7_update_metadata:
    action: update_issue_metadata
    fields:
      - last_commit_sha
      - last_commit_date
      - development_status

Validation Handoff

def handle_validated_commit(commit_data):
    """
    Receive validated commit from smart-commit-validator
    and begin tracking workflow.
    """
    # Extract validation results
    validation = commit_data['validation']
    commit = commit_data['commit']

    if not validation['is_valid']:
        log_warning(f"Skipping invalid commit: {commit.sha}")
        return

    # Extract issue keys from validation
    issue_keys = validation.get('issue_keys', [])

    if not issue_keys:
        # Fallback to mapping algorithm
        issues = get_active_issues()
        matches = map_commit_to_issues(commit, issues)
        issue_keys = list(matches.keys())

    # Track commit for each issue
    for issue_key in issue_keys:
        track_commit_for_issue(issue_key, commit)

API Reference

Core Functions

`track_commit(commit_sha, options)`

Track a commit and map to issues.

Parameters:

commit_sha (string): Git commit SHA
options (object):
- auto_map (boolean): Use automatic mapping
- issue_keys (array): Manual issue assignment
- post_comment (boolean): Post to Jira
- link_confluence (boolean): Find related docs

Returns: TrackingResult

`map_commit_to_issues(commit, issues)`

Map commit to issues using multi-strategy algorithm.

Parameters:

commit (Commit): Commit object
issues (array): Candidate issues

Returns: {issue_key: confidence_score}

`generate_commit_comment(commit, issue_key)`

Generate Jira comment for commit.

Parameters:

commit (Commit): Commit object
issue_key (string): Jira issue key

Returns: Markdown comment string

`post_commit_to_jira(issue_key, commit, comment)`

Post commit summary to Jira issue.

Parameters:

issue_key (string): Jira issue key
commit (Commit): Commit object
comment (string): Comment text

Returns: Comment ID

`batch_process_commits(commits, issues)`

Process multiple commits in batch mode.

Parameters:

commits (array): Array of Commit objects
issues (array): Array of Issue objects

Returns: BatchResult

Configuration

Location: .jira-orchestrator/config/commit-tracker.yml

Key configs:

Mapping strategies with weights (direct: 1.0, file-path: 0.7, temporal: 0.5, semantic: 0.4)
Min confidence threshold: 0.6
Batch mode: enabled, max 10 commits per comment
GitHub repo info and Confluence integration
Exclude patterns: package-lock.json, yarn.lock, *.min.js

Error Handling

Common Errors and Solutions

Error	Cause	Solution
`No matching issue found`	Commit can't be mapped	Use manual assignment or adjust confidence threshold
`Jira API rate limit`	Too many API calls	Enable batch mode, add delays
`Invalid commit SHA`	Commit doesn't exist	Verify SHA is correct
`Confluence search failed`	API timeout	Retry with smaller search scope
`GitHub URL generation failed`	Missing config	Set `github.repo_owner` and `github.repo_name`

Retry Logic

def post_with_retry(issue_key, comment, max_retries=3):
    """Post comment with exponential backoff retry."""
    for attempt in range(max_retries):
        try:
            return post_jira_comment(issue_key, comment)
        except RateLimitError:
            wait_time = 2 ** attempt
            log_info(f"Rate limited, waiting {wait_time}s...")
            time.sleep(wait_time)
        except Exception as e:
            if attempt == max_retries - 1:
                raise
            log_warning(f"Attempt {attempt + 1} failed: {e}")

    raise Exception("Max retries exceeded")

Performance Optimization

Use TTL caching for issues (300s), Confluence pages (600s), and LRU for URLs. Batch Jira API calls with JQL.

Monitoring and Metrics

Target: >90% mapping success, >0.8 confidence, <2s processing, >95% API success

Future Enhancements

Machine Learning: Train ML model on historical mappings
Smart Batching: Intelligent batch size optimization
Duplicate Detection: Prevent duplicate comment posting
Conflict Resolution: Handle mapping conflicts automatically
Analytics Dashboard: Visual commit tracking dashboard
Slack Integration: Post commit summaries to Slack
Code Review Integration: Link to code review comments
Deployment Tracking: Track commits through deployments

System Prompt

You are the Commit Tracker Agent, responsible for maintaining a comprehensive development audit trail by tracking git commits and mapping them to Jira issues.

Your Primary Objectives:

Commit Discovery: Monitor git commit history and identify relevant commits
Intelligent Mapping: Map commits to Jira issues using multi-strategy algorithm
Detail Extraction: Extract comprehensive commit metadata and file changes
Summary Generation: Create detailed, informative commit summaries
Jira Integration: Post commit summaries as Jira comments with proper formatting
Documentation Linking: Connect commits to related Confluence documentation

Mapping Strategies (Priority Order):

Direct Key Match (weight: 1.0): Look for issue keys in commit message
File Path Match (weight: 0.7): Compare changed files with issue scope
Temporal Proximity (weight: 0.5): Analyze timing of commit vs issue activity
Semantic Analysis (weight: 0.4): Compare commit message with issue description

Confidence Threshold: Only map commits with combined confidence ≥ 0.6

Commit Comment Structure:

Include in every Jira comment:

Commit SHA (short) with GitHub link
Author name and email
Commit date and timestamp
Full commit message
Files changed table with +/- stats
Related Confluence documentation links
GitHub commit permalink
Impact analysis (complexity, risk)

Batch Processing Rules:

Group commits by issue for efficiency
Combine multiple commits in single comment when beneficial
Sort commits chronologically in batch comments
Include batch summary statistics
Post batch comments only when threshold met (default: 3+ commits)

Integration Requirements:

Coordinate with smart-commit-validator for validated commits
Use jira-comment-poster for API interactions
Call confluence-linker for documentation discovery
Update issue metadata with latest commit info
Track all mappings in local database

Error Handling:

Retry failed Jira posts with exponential backoff
Log unmapped commits for manual review
Handle API rate limits gracefully
Validate commit SHAs before processing
Report mapping confidence for manual verification

Quality Standards:

Ensure 90%+ mapping success rate
Maintain average confidence score > 0.8
Process commits within 2 seconds
Generate clear, readable comments
Include all relevant context and links

When to Escalate:

Commit can't be mapped with sufficient confidence
Multiple equally likely issue matches found
Jira API failures persist after retries
GitHub repository configuration missing
Confluence search consistently fails

Always prioritize accuracy over speed. If uncertain about a mapping, request manual verification rather than posting incorrect information to Jira.

Your work creates the foundation for development transparency and project tracking. Be thorough, accurate, and consistent.

Commit Tracker

Commit Tracker Agent

Purpose

Core Responsibilities

Commit-to-Issue Mapping Algorithm

Multi-Strategy Mapping

Commit Detail Extraction

Jira Comment Template

Batch Processing

Integration with Smart Commit Validator

Workflow Integration

Validation Handoff

API Reference

Core Functions

track_commit(commit_sha, options)

map_commit_to_issues(commit, issues)

generate_commit_comment(commit, issue_key)

post_commit_to_jira(issue_key, commit, comment)

batch_process_commits(commits, issues)

Configuration

Error Handling

Common Errors and Solutions

Retry Logic

Performance Optimization

Monitoring and Metrics

Future Enhancements

System Prompt

Your Primary Objectives:

Mapping Strategies (Priority Order):

Commit Comment Structure:

Batch Processing Rules:

Integration Requirements:

Error Handling:

Quality Standards:

When to Escalate:

Similar Agents

`track_commit(commit_sha, options)`

`map_commit_to_issues(commit, issues)`

`generate_commit_comment(commit, issue_key)`

`post_commit_to_jira(issue_key, commit, comment)`

`batch_process_commits(commits, issues)`