desloppify-scan
Run full codebase quality scan (linters + subjective review), persist state, compute scores.
- License 许可
- MIT
- 最近更新
- 2026-10-09
- Stars
- ★ 0
- Forks
- 0
01 前言 Front Matter
- name
- desloppify-scan
- triggers
- [ "user" ]
- description
- Run full codebase quality scan (linters + subjective review), persist state, compute scores.
03 SKILL.md 正文
Follow ./reference/workflows/scan.md.
For Step 3 (Subjective Review), execute the 4 batches sequentially in this session — Devin's child sessions run on separate VMs and can't share .desloppify/state.json, so the parallel-subagent pattern is flattened to a single thorough pass. Work through the batches in order (Architecture → Correctness → Code Quality → External), following the per-batch instructions in ./reference/workflows/review.md for each. Accumulate the four JSON outputs, then continue with Step 4.
04 同仓库更多 Skills
desloppify-triage
Analyze, cluster, and plan the work queue from scan findings.
desloppify-status
Show current desloppify scores and progress.
desloppify-review
Run subjective review only (sequential passes scoring 20 quality dimensions).
desloppify-resolve
Mark work items as resolved.
desloppify-next
Show and work the next item in the triage queue.