@hopla/claude-setup 1.1.1 → 1.1.2
This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
|
@@ -33,32 +33,41 @@ Work through each task in the plan sequentially. For each task:
|
|
|
33
33
|
2. **Follow the pattern** referenced in the plan — do not invent new patterns
|
|
34
34
|
3. **Implement** only what the task specifies — nothing more
|
|
35
35
|
4. **Validate** the task using the method specified in the plan's validate field
|
|
36
|
-
5. **
|
|
36
|
+
5. **Report completion** with a brief status: what was done, what was skipped, any decision made
|
|
37
|
+
6. **Do not proceed** to the next task if the current one fails validation
|
|
37
38
|
|
|
38
|
-
|
|
39
|
+
**Trust but Verify — pause and report if:**
|
|
40
|
+
- A file referenced in the plan doesn't exist
|
|
41
|
+
- The actual pattern in the code differs from what the plan assumed
|
|
42
|
+
- A task is ambiguous or has multiple valid implementations
|
|
43
|
+
- Something unexpected is discovered that could affect subsequent tasks
|
|
44
|
+
|
|
45
|
+
Do not improvise silently. When in doubt, stop and ask.
|
|
39
46
|
|
|
40
47
|
## Step 4: Run Full Validation Pyramid
|
|
41
48
|
|
|
42
49
|
After all tasks are complete, run the full validation sequence in order.
|
|
43
50
|
**Do not skip levels. Do not proceed if a level fails.**
|
|
44
51
|
|
|
52
|
+
Use the exact commands from the plan's **Validation Checklist**. If not specified, read `CLAUDE.md` section "Development Commands" to find the correct commands.
|
|
53
|
+
|
|
45
54
|
### Level 1 — Lint & Format
|
|
46
|
-
Run the project's lint and format check.
|
|
55
|
+
Run the project's lint and format check (e.g. `npm run lint`, `uv run ruff check .`).
|
|
47
56
|
Fix any issues before continuing.
|
|
48
57
|
|
|
49
58
|
### Level 2 — Type Check
|
|
50
|
-
Run the project's type checker.
|
|
59
|
+
Run the project's type checker (e.g. `npm run type-check`, `uv run mypy .`).
|
|
51
60
|
Fix all type errors before continuing.
|
|
52
61
|
|
|
53
62
|
### Level 3 — Unit Tests
|
|
54
|
-
Run the project's unit test suite.
|
|
63
|
+
Run the project's unit test suite (e.g. `npm run test`, `uv run pytest`).
|
|
55
64
|
If tests fail:
|
|
56
65
|
- Investigate the root cause
|
|
57
66
|
- Fix the code (not the tests)
|
|
58
67
|
- Re-run until all pass
|
|
59
68
|
|
|
60
69
|
### Level 4 — Integration Tests
|
|
61
|
-
Run integration tests or manual verification as specified in the plan.
|
|
70
|
+
Run integration tests or manual verification as specified in the plan (e.g. `npm run test:e2e`, manual curl).
|
|
62
71
|
Verify the feature works end-to-end.
|
|
63
72
|
|
|
64
73
|
### Level 5 — Human Review (flag for user)
|
|
@@ -125,7 +125,28 @@ Add to `.gitignore` (create if it doesn't exist):
|
|
|
125
125
|
.agents/system-reviews/
|
|
126
126
|
```
|
|
127
127
|
|
|
128
|
-
## Step 5:
|
|
128
|
+
## Step 5: Create .claude/commands/ (optional but recommended)
|
|
129
|
+
|
|
130
|
+
Create `.claude/commands/` at the project root for project-specific commands that override or extend the global ones.
|
|
131
|
+
|
|
132
|
+
Common project-specific commands to create:
|
|
133
|
+
|
|
134
|
+
**`validate.md`** — runs the full validation sequence for this project:
|
|
135
|
+
```markdown
|
|
136
|
+
---
|
|
137
|
+
description: Run full validation for this project
|
|
138
|
+
---
|
|
139
|
+
Run in order, stop if any level fails:
|
|
140
|
+
1. `[lint command]`
|
|
141
|
+
2. `[type check command]`
|
|
142
|
+
3. `[test command]`
|
|
143
|
+
```
|
|
144
|
+
|
|
145
|
+
Ask the user: "Do you want me to create a project-specific `/validate` command with the commands from your stack?"
|
|
146
|
+
|
|
147
|
+
If yes, create `.claude/commands/validate.md` using the dev commands collected in Topic E.
|
|
148
|
+
|
|
149
|
+
## Step 6: Confirm and Save
|
|
129
150
|
|
|
130
151
|
Show the draft `CLAUDE.md` to the user and ask:
|
|
131
152
|
> "Does this accurately reflect the project's rules? Any corrections before I save it?"
|
|
@@ -123,8 +123,8 @@ Run in this order — do not proceed if a level fails:
|
|
|
123
123
|
Before finishing, review the plan against these criteria:
|
|
124
124
|
|
|
125
125
|
- [ ] Every task has a specific file path (no vague "update the component")
|
|
126
|
-
- [ ] Every task
|
|
127
|
-
- [ ] Validation commands
|
|
126
|
+
- [ ] Every task has: Action, File, Pattern, Details, Gotcha, Validate — no field left empty
|
|
127
|
+
- [ ] Validation Checklist has **exact commands** from `CLAUDE.md` or `package.json` (not vague "run lint")
|
|
128
128
|
- [ ] The plan is complete enough that another agent can execute it without this conversation
|
|
129
129
|
- [ ] No ambiguous requirements left unresolved
|
|
130
130
|
|