Jira Sync - Synchronize Jira Content

Synchronize Jira issues and comments from configured projects into the AI Memory jira-data collection.

Usage

# Incremental sync (default) - only fetch updated issues
/aim-jira-sync

# Full sync - fetch all issues and comments
/aim-jira-sync --full

# Sync specific project
/aim-jira-sync --project BMAD

# Check sync status (last sync time, items synced, errors)
/aim-jira-sync --status

# Full sync for specific project
/aim-jira-sync --full --project BMAD

Options

• --incremental - Sync only updated issues since last sync (default)
• --full - Full sync: fetch all issues and comments from scratch
• --project <key> - Sync specific project (default: all configured projects)
• --status - Display sync status (last sync time, items synced, errors)

Sync Modes

Incremental Sync (Default)

• Fetches issues updated since last sync timestamp
• Faster and more efficient for regular updates
• Uses JQL: project = PROJ AND updated >= "2026-02-01 00:00"
• Updates existing documents in-place (by jira_comment_id for comments)

Full Sync

• Fetches all issues and comments from scratch
• Use for initial setup or after schema changes
• Warning: Can be slow for large projects (hundreds/thousands of issues)
• Overwrites existing documents

Configuration

Sync requires Jira credentials in .env:

JIRA_INSTANCE_URL=https://company.atlassian.net
[email protected]
JIRA_API_TOKEN=your_api_token_here
JIRA_PROJECTS=BMAD,PROJ,DEV
JIRA_SYNC_ENABLED=true
JIRA_SYNC_DELAY_MS=100

Examples

# Daily incremental sync (recommended)
/aim-jira-sync

# Initial full sync for new project
/aim-jira-sync --full --project NEWPROJ

# Check last sync status
/aim-jira-sync --status

# Full refresh of all projects (use sparingly)
/aim-jira-sync --full

Implementation Reference

This skill invokes scripts/jira_sync.py:

cd /path/to/ai-memory
python3 scripts/jira_sync.py --incremental
python3 scripts/jira_sync.py --full --project BMAD
python3 scripts/jira_sync.py --status

Technical Details

• Rate Limiting: Configurable delay between API requests (default 100ms)
• Pagination: Token-based for issues, offset-based for comments
• Deduplication: SHA256 content hashing prevents duplicate storage
• Document Format: ADF (Atlassian Document Format) → plain text
• Collection: jira-data (separate from code-patterns/conventions/discussions)
• Tenant Isolation: group_id = Jira instance hostname

Notes

• Requires Jira Cloud API token (create at id.atlassian.com)
• Sync logs written to ~/.ai-memory/logs/activity.log
• First sync can take several minutes for large projects
• Incremental sync timestamp stored in jira_sync_state.json
• Comments are linked to parent issues via jira_issue_key