/tidy

Automated backlog cleanup. Non-interactive.

What This Does

1. Lint all open issues
2. Enrich issues scoring < 70
3. Deduplicate similar issues
4. Close stale issues (>90 days untouched, not blocked)
5. Migrate legacy labels
6. Report what changed

Process

1. Lint All Issues

Run /issue lint --all to score every open issue.

2. Enrich Failures

For each issue scoring < 70:

• Run /issue enrich to fill gaps
• Re-lint to confirm improvement
• Skip issues scoring < 30 (need human input, not enrichment)

3. Deduplicate

Compare issue titles and bodies for semantic similarity:

• Same domain label + similar problem statement = potential duplicate
• Present duplicates for confirmation before closing
• Close the less-detailed duplicate with "Duplicate of #N"

4. Close Stale

# Issues untouched for 90+ days without blocked label
gh issue list --state open --json number,title,updatedAt,labels | \
  jq '[.[] | select(
    ((.updatedAt | fromdateiso8601) < (now - 7776000)) and
    (.labels | map(.name) | index("blocked") | not)
  )]'

For each stale issue:

gh issue close N --comment "Closing as stale (>90 days without update). Reopen if still relevant."

5. Migrate Legacy Labels

# Find issues with legacy labels
gh issue list --state open --label "priority/p0" --json number
gh issue list --state open --label "P0" --json number
gh issue list --state open --label "type/bug" --json number

Migrate each:

gh issue edit N --remove-label "priority/p0" --add-label "p0"

6. Report

TIDY REPORT
===========

Issues linted: 20
Issues enriched: 3 (#35: 45→82, #38: 55→78, #41: 62→75)
Duplicates closed: 1 (#39 → duplicate of #42)
Stale closed: 2 (#12, #15)
Labels migrated: 1 (#18: priority/p0 → p0)

Backlog health: 60% → 78% ready for execution

Related

• /issue lint — Per-issue scoring
• /issue enrich — Gap-filling with research
• /backlog — Read-only health dashboard
• /groom — Full interactive grooming (tidy is the quick non-interactive version)