| name | trellis-check |
| description | Comprehensive quality verification: spec compliance, lint, type-check, tests, cross-layer data flow, code reuse, and consistency checks. Use when code is written and needs quality verification, before committing changes, or to catch context drift during long sessions. |
Code Quality Check
Comprehensive quality verification for recently written code. Combines spec compliance, cross-layer safety, and pre-commit checks.
Step 1: Identify What Changed
git diff --name-only HEAD
git status
Step 2: Read Applicable Specs
python3 ./.trellis/scripts/get_context.py --mode packages
For each changed package/layer, read the spec index and follow its Quality Check section:
cat .trellis/spec/<package>/<layer>/index.md
Read the specific guideline files referenced — the index is a pointer, not the goal.
Step 3: Run Project Checks
Run the project's lint, type-check, and test commands. Fix any failures before proceeding.
Step 4: Review Against Checklist
Code Quality
Test Coverage
Spec Sync
"If I fixed a bug or discovered something non-obvious, should I document it so future me won't hit the same issue?" → If YES, update the relevant spec doc.
Step 5: PRD Acceptance Criteria Verification
5A. Locate the PRD
Find the current task's prd.md. Use one of these methods in order:
- Run
python3 ./.trellis/scripts/task.py current --source to get the active task directory
- If no active task, check for a task directory matching the current git branch name under
.trellis/tasks/
- If neither works, skip this step and note that the PRD could not be located
5B. Parse Acceptance Criteria
Read the ## Acceptance Criteria section from the PRD. Parse all checkbox items:
- [ ] <description> — unchecked, needs verification- [x] <description> — already checked, skip
5C. Classify Each Unchecked Item
For each - [ ] item, determine whether it can be auto-verified by this agent:
Auto-verifiable (this agent can confirm directly):
- Lint passes, type-check passes, tests pass
- A specific file exists at a known path
- A specific code pattern is present (grep for function signature, import, class name, etc.)
- A specific configuration value is set
- A command produces expected output
Manual-only (requires the user, specific hardware, or subjective judgment):
- Requires a specific hardware device (e.g., "Pi 上测试唤醒词检测成功率 > 90%")
- Requires manual UI interaction (e.g., "手动测试按钮点击效果")
- Requires subjective human judgment (e.g., "代码可读性良好")
- Requires an environment this agent cannot access (e.g., "在 Windows 上验证")
- Contains vague or qualitative criteria that cannot be checked programmatically
5D. Verify and Check Off
For each auto-verifiable item that passes: use the Edit tool to change - [ ] to - [x] in the PRD.
For each auto-verifiable item that fails: leave as - [ ] and note the failure in the report.
For each manual-only item: leave as - [ ] and note in the report why it cannot be auto-verified.
5E. Report
Report a summary in the check output:
## PRD Acceptance Criteria Verification
### Verified and checked off
- [x] Linter passes
- [x] Type checker passes
### Failed (auto-verifiable, needs fixing)
- [ ] Tests pass → 3 test failures in test_auth.py
### Skipped (manual verification required)
- [ ] Pi 上测试唤醒词检测成功率 > 90% → 需要 Pi 硬件
- [ ] 手动测试 UI 交互 → 需要手动操作
Do NOT check off items you cannot verify. A false [x] is worse than a missing [x].
Step 6: Cross-Layer Dimensions (if applicable)
Skip this step if your change is confined to a single layer.
A. Data Flow (changes touch 3+ layers)
B. Code Reuse (modifying constants, creating utilities)
C. Import/Dependency (creating new files)
D. Same-Layer Consistency
Step 7: Report and Fix
Report violations found and fix them directly. Re-run project checks after fixes.
