fetch-data

Flow Fetch Data

Collect flow data from the configured source for the current project.

Steps

1. Detect git remote and resolve project config:

   git remote get-url origin
   

   Look up config in ~/.tommymorgan/flows/ by matching the remote URL. If no config found, report: "No flow configuration found for remote URL: . Run /flow:setup to configure." and stop. If config file is invalid JSON, report: "Configuration file is invalid: . Run /flow:setup to reconfigure." and stop.

2. Check authentication for the configured source:

   • GitHub: gh auth status
   • If not authenticated, report the error with remediation steps and stop.

3. Read existing flow data (if any) from ~/.tommymorgan/flows/<slug>-data.json

4. Determine the since date:

   • If existing data exists, use the most recent finishedAt timestamp
   • If --all flag provided, ignore existing data and fetch everything
   • If no existing data, fetch all

5. Fetch work items using the configured adapter:

   • GitHub: For each configured repo, fetch merged PRs via gh api:
     • Get PR list with pagination
     • For each merged PR, fetch commits to get the earliest commit authored date
     • If configured start boundary data is unavailable, fall back to PR created_at and include a warning
     • Calculate Cycle Time: finished - started + 1 (calendar days, UTC)
   • Linear / Git: Not yet implemented — report "Adapter not yet available" and stop

6. Merge new items with existing data (deduplicate by id + source)

7. Write updated data to ~/.tommymorgan/flows/<slug>-data.json atomically

8. Report results:

   Flow data updated.
   - Total items: {count}
   - New items fetched: {new_count}
   - Date range: {earliest_date} to {latest_date}
   - Source: {source} ({repos})
   {warnings if any fallbacks occurred}
   

Arguments

• --all — Re-fetch all data (ignore incremental)
• --repo owner/repo — Override configured repo(s)