task-orchestrator: Command for Claude Code

Check the schema version of the Task Orchestrator database on a Docker volume.

Instructions

1. Parse the volume name from the user's command, defaulting to "mcp-task-data" if not provided
2. Create a temporary SQL file with the query
3. Run the Docker command mounting both the volume and the project directory
4. Display the results in a clear, formatted way
5. Clean up the temporary SQL file

Working Pattern (Windows Git Bash Compatible)

Due to shell escaping issues on Windows, use a mounted SQL file approach:

Step 1: Create temporary SQL file in project root:

-- File: .tmp_check_schema.sql (in project root directory)
.mode box
.headers on

SELECT
    installed_rank,
    version,
    description,
    type,
    script,
    installed_on,
    execution_time,
    success
FROM flyway_schema_history
ORDER BY installed_rank DESC
LIMIT 10;

Temp File Location: Create .tmp_check_schema.sql in the project root directory (D:\Projects\task-orchestrator)

Step 2: Run via Docker with mounted SQL file:

powershell.exe -Command "docker run --rm -v {{VOLUME_NAME}}:/data -v '${PWD}:/work' alpine sh -c 'apk add --no-cache sqlite >/dev/null 2>&1 && sqlite3 /data/current-tasks.db < /work/.tmp_check_schema.sql'"

Step 3: Clean up:

rm .tmp_check_schema.sql

Note: The alpine/sqlite image doesn't work well with inline SQL on Windows. Use alpine with sqlite installed instead.

Expected Output Format

Present the results with:

• Current schema version (the most recent migration)
• List of recent migrations (up to 10)
• Success status for each migration
• Installation timestamps

If the query fails:

• Explain that the volume may not exist, or
• The database may not have been initialized with Flyway, or
• The database file doesn't exist yet

Additional Helpful Commands

After showing the results, offer these follow-up options:

1. Show all migrations: Remove the LIMIT from the SQL query
2. Show table structure: Create SQL file with .schema flyway_schema_history
3. Show all tables: Create SQL file with .tables
4. Check database file:

   docker run --rm -v {{VOLUME_NAME}}:/data alpine ls -lh /data/