You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
When invoked, generate a markdown file containing the following template structure:
22
22
23
23
```markdown
24
-
# Feature: [SHORTDESC] ([ISSUE_NUMBER])
24
+
---
25
+
task: [SHORTDESC] (Feature #[ISSUE_NUMBER])
26
+
test_command: "[REPLACE: command to verify feature works]"
27
+
---
25
28
26
-
> **⚠️ IMPORTANT**: Before implementing this feature, READ `/CLAUDE.md` first.
29
+
# Task: [SHORTDESC] (Feature #[ISSUE_NUMBER])
27
30
28
-
## Summary
31
+
> **⚠️ IMPORTANT**: Before implementing this feature, READ `/CLAUDE.md` first.
29
32
30
33
_A concise summary of the feature and its purpose._
31
34
32
-
## User Stories
33
-
34
-
- As a [user type], I want [goal] so that [benefit].
35
-
36
-
## Acceptance Criteria
37
-
38
-
-[ ]
39
-
-[ ]
40
-
-[ ]
41
-
42
-
## Technical Requirements
35
+
## Requirements
43
36
44
-
_Key technical considerations, constraints, or dependencies._
37
+
1.[Requirement 1]
38
+
2.[Requirement 2]
39
+
3.[Requirement 3]
45
40
46
-
## Implementation Notes
41
+
## Success Criteria
47
42
48
-
_High-level approach or architecture decisions._
43
+
1.[ ][Testable criterion 1]
44
+
2.[ ][Testable criterion 2]
45
+
3.[ ][Testable criterion 3]
46
+
4.[ ][Testable criterion 4]
47
+
5.[ ][Testable criterion 5]
48
+
6.[ ][Testable criterion 6]
49
49
50
-
## Dependencies
50
+
## Example Output
51
51
52
-
-_List any blocking issues, APIs, or services._
53
-
54
-
## Out of Scope
55
-
56
-
_What is explicitly NOT part of this feature._
57
-
58
-
## Success Metrics
59
-
60
-
_How will we measure success?_
61
-
62
-
## Additional Context
63
-
64
-
_Mockups, references, or related discussion._
52
+
\`\`\`
53
+
[Expected output or behavior demonstration]
54
+
\`\`\`
65
55
66
56
---
67
57
68
-
## Completion
58
+
## Ralph Instructions
69
59
70
-
Output `<promise>DONE</promise>` when all tests green. --max-iterations 50 --completion-promise "DONE"
60
+
1. Work on the next incomplete criterion (marked [])
61
+
2. Check off completed criteria (change [] to [x])
62
+
3. Run tests after changes
63
+
4. Commit your changes frequently
64
+
5. When ALL criteria are [x], output: `<ralph>COMPLETE</ralph>`
65
+
6. If stuck on the same issue 3+ times, output: `<ralph>GUTTER</ralph>`
71
66
```
72
67
73
68
---
74
69
75
70
## Workflow
76
71
77
72
1._PARSE_ input into `ISSUE_NUMBER` and `SHORTDESC`
78
-
2._GENERATE_ a new markdown file at `.claude/specs/feature-[ISSUE_NUMBER]-[SHORTDESC]/SPEC.md` containing the template above (substitute in the values)
73
+
2._GENERATE_ a new markdown file at `.claude/specs/feat-[ISSUE_NUMBER]-[SHORTDESC]/SPEC.md` containing the template above (substitute in the values)
79
74
3._CONFIRM_ file creation and path
80
75
81
76
---
@@ -87,33 +82,28 @@ Before the feature spec is considered complete and ready for implementation, ver
87
82
### Required Fields
88
83
89
84
-[ ]**Summary** clearly articulates what the feature does and why it matters
90
-
-[ ]**User Stories** has at least one complete user story (As a [user], I want [goal], so that [benefit])
91
-
-[ ]**Acceptance Criteria** has 3+ testable, unambiguous criteria (checkbox format)
92
-
-[ ]**Technical Requirements** lists key constraints, APIs, or architectural decisions
93
-
94
-
### Scope Definition
95
-
96
-
-[ ]**Out of Scope** explicitly states what is NOT part of this feature
97
-
-[ ]**Dependencies** lists any blocking issues, external services, or prerequisite work
85
+
-[ ]**Requirements** has at least 3 specific implementation requirements
0 commit comments