Slop Fix - Auto-Fix Code Quality Issues
Applies deterministic fixes to high-confidence findings. Run slop_scan or slop_scan_now first.
Inputs
| Input | Type | Default | Purpose |
|---|---|---|---|
dry_run |
bool | false | Preview changes without applying |
min_confidence |
float | 0.90 | Minimum confidence (0.0-1.0) |
limit |
int | 20 | Max findings to fix per run |
commit |
bool | true | Commit changes to git |
Workflow
1. Check Database
- Verify
~/.config/aa-workflow/slop.dbexists - Query:
SELECT COUNT(*) FROM findings WHERE status = 'open' - If no DB or 0 findings: output "No open findings" and skip
2. Get Fixable Findings
- Call
slop_fixableMCP tool (if available) withmin_confidence,limit - Or query DB for findings with confidence >= min_confidence
3. Apply Fixes
- Call
slop_fixMCP tool (if available) with:dry_run,min_confidence,limit,commit
- Or run slop fix CLI if tool not available
4. Log Session
memory_session_log("Slop auto-fix completed", "Dry run" or "Applied fixes")
5. Output Format
## Slop Auto-Fix Results
### Database Status
<db_status>
### Fixable Findings
<fixable_list>
### Fix Results
<fix_result>
---
Run with `dry_run=true` to preview changes before applying.
Fix Types (confidence >90%)
- unused_imports: Remove import lines
- unused_variables: Remove or prefix with
_ - bare_except: Change to
except Exception: - empty_except: Add
logger.exception() - dead_code: Remove unused functions/classes
Key Details
- Depends on:
slop_scanorslop_scan_now - Chains to:
create_mr,review_local_changes - Always run scan before fix to ensure fresh findings