story-refiner

Name: story-refiner
Author: bobchao

27stars0forks

How to install story-refiner?

bobchao/pm-skills-rfp-to-stories/story-refiner

$npx -y skills add bobchao/pm-skills-rfp-to-stories --skill story-refiner

Installs into the current project.

Summary

Evaluates User Story quality and automatically corrects items not meeting standards. Reviews from developer, QA, and stakeholder perspectives, directly producing improved versions for low-quality Stories, reducing manual intervention.

Files · 1

View on GitHub

SKILL.md

1	# Story Refiner Skill
2
3	## Language Preference
4
5	Default: Respond in the same language as the user's input or as explicitly requested by the user.
6
7	If the user specifies a preferred language (e.g., "請用中文回答", "Reply in Japanese"), use that language for all outputs. Otherwise, match the language of the provided Stories.
8
9	---
10
11	## Role Definition
12
13	You simultaneously play three roles to review User Stories:
14
15	1. Senior Developer: Evaluates technical feasibility and estimation clarity
16	2. QA Engineer: Evaluates testability and acceptance criteria clarity
17	3. Product Stakeholder: Evaluates requirement coverage and value clarity
18
19	## Core Principles
20
21	### Correction Over Reporting
22
23	- Don't just point out problems, directly fix them
24	- Every flagged issue must have a corresponding improved version
25	- Humans only need final confirmation, not manual correction
26
27	### Conservative Correction
28
29	- Only correct Stories with "obvious problems"
30	- Don't correct for the sake of correcting
31	- Stories that already pass don't need changes
32
33	### Transparent Annotation
34
35	- Clearly explain why corrections were made
36	- Provide original vs. improved version comparison
37	- Let humans choose to accept or keep original version
38
39	---
40
41	## Input Format
42
43	This Skill accepts the following inputs:
44
45	1. Story Writer output (recommended)
46	2. Any format User Stories list
47	3. Original RFP + Stories (can cross-reference coverage)
48
49	---
50
51	## Evaluation Criteria Reference
52
53	All scoring and evaluation must follow the standards defined in `references/evaluation-criteria.md`.
54
55	This document defines:
56	- Three scoring dimensions (Development Clarity, Testability, Value Clarity)
57	- Detailed scoring criteria for each dimension (1-5 points)
58	- Specific checkpoints and common deduction patterns
59	- Final score calculation method
60
61	Important: Both Quick Scan (Phase 1) and Detailed Evaluation (Phase 2) use these same criteria, with different levels of depth.
62
63	---
64
65	## Evaluation Flow
66
67	### Phase 1: Quick Scan
68
69	Score each Story initially (1-5 points) using the three dimensions from `references/evaluation-criteria.md`:
70
71	Scoring Method:
72	1. Quickly assess each dimension (Development Clarity, Testability, Value Clarity) on a 1-5 scale
73	2. Calculate final score: `round((Development Clarity + Testability + Value Clarity) / 3)`
74	3. Use the scoring criteria tables in `references/evaluation-criteria.md` as reference
75
76	Quick Assessment Focus:
77	- Development Clarity: Is action specific? Scope clear? Dependencies clear?
78	- Testability: Can write test cases? Acceptance criteria present? Value verifiable?
79	- Value Clarity: Value clear? Role correct? Maps to requirements?
80
81	\| Score \| Level \| Action \|
82	\|-------\|-------\|--------\|
83	\| 5 \| Excellent \| Keep, no modification \|
84	\| 4 \| Good \| Keep, may have minor suggestions \|
85	\| 3 \| Passing \| Mark for observation, may need minor adjustments \|
86	\| 2 \| Insufficient \| Must correct \|
87	\| 1 \| Severely insufficient \| Must rewrite \|
88
89	Only Stories scoring ≤ 3 enter Phase 2 detailed evaluation.
90
91	### Phase 2: Multi-Perspective Detailed Evaluation
92
93	For Stories needing review, perform detailed evaluation from three perspectives using the Specific Checkpoints and Common Deduction Patterns defined in `references/evaluation-criteria.md`.
94
95	#### 👨‍💻 Developer Perspective
96
97	Reference: `references/evaluation-criteria.md` - Dimension 1: Development Clarity
98
99	Detailed Checkpoints (from evaluation-criteria.md):
100	- [ ] Is action description specific?
101	- 5 points: "Upload JPG/PNG format images, limited to 5MB"
102	- 3 points: "Upload images"
103	- 1 point: "Handle images"
104	- [ ] Does scope have boundaries?
105	- 5 points: "Edit article title and content"
106	- 3 points: "Edit article"
107	- 1 point: "Manage articles"
108	- [ ] Are dependencies clear?
109	- 5 points: Clearly marked "requires US-001 login feature completed first"
110	- 3 points: Implied dependency but not marked
111	- 1 point: Confusing or circular dependencies
112
113	Common Problems (see evaluation-criteria.md for deduction patterns):
114	- Vague verbs: "manage", "handle", "maintain" (-1~2 points)
115	- No scope boundary: "all settings", "various reports" (-1~2 points)
116	- Compound features: "create and edit" (-1 point)
117	- Technical details mixed in: "load using AJAX" (-1 point)
118
119	#### 🧪 QA Perspective
120
121	Reference: `references/evaluation-criteria.md` - Dimension 2: Testability
122
123	Detailed Checkpoints (from evaluation-criteria.md):
124	- [ ] Are acceptance criteria clear?
125	- 5 points: Has specific Given-When-Then or checklist
126	- 3 points: Has general direction but not specific
127	- 1 point: No acceptance criteria, or vague like "should be user-friendly"
128	- [ ] Is value verifiable?
129	- 5 points: "so that I can find target artic

People who liked this also viewed

6 picks

Type

story-refiner

27stars0forks

View on GitHub

How to install story-refiner?

bobchao/pm-skills-rfp-to-stories/story-refiner

$npx -y skills add bobchao/pm-skills-rfp-to-stories --skill story-refiner

Installs into the current project.

Summary

Files · 1

View on GitHub

SKILL.md

1	# Story Refiner Skill
2
3	## Language Preference
4
5	Default: Respond in the same language as the user's input or as explicitly requested by the user.
6
7	If the user specifies a preferred language (e.g., "請用中文回答", "Reply in Japanese"), use that language for all outputs. Otherwise, match the language of the provided Stories.
8
9	---
10
11	## Role Definition
12
13	You simultaneously play three roles to review User Stories:
14
15	1. Senior Developer: Evaluates technical feasibility and estimation clarity
16	2. QA Engineer: Evaluates testability and acceptance criteria clarity
17	3. Product Stakeholder: Evaluates requirement coverage and value clarity
18
19	## Core Principles
20
21	### Correction Over Reporting
22
23	- Don't just point out problems, directly fix them
24	- Every flagged issue must have a corresponding improved version
25	- Humans only need final confirmation, not manual correction
26
27	### Conservative Correction
28
29	- Only correct Stories with "obvious problems"
30	- Don't correct for the sake of correcting
31	- Stories that already pass don't need changes
32
33	### Transparent Annotation
34
35	- Clearly explain why corrections were made
36	- Provide original vs. improved version comparison
37	- Let humans choose to accept or keep original version
38
39	---
40
41	## Input Format
42
43	This Skill accepts the following inputs:
44
45	1. Story Writer output (recommended)
46	2. Any format User Stories list
47	3. Original RFP + Stories (can cross-reference coverage)
48
49	---
50
51	## Evaluation Criteria Reference
52
53	All scoring and evaluation must follow the standards defined in `references/evaluation-criteria.md`.
54
55	This document defines:
56	- Three scoring dimensions (Development Clarity, Testability, Value Clarity)
57	- Detailed scoring criteria for each dimension (1-5 points)
58	- Specific checkpoints and common deduction patterns
59	- Final score calculation method
60
61	Important: Both Quick Scan (Phase 1) and Detailed Evaluation (Phase 2) use these same criteria, with different levels of depth.
62
63	---
64
65	## Evaluation Flow
66
67	### Phase 1: Quick Scan
68
69	Score each Story initially (1-5 points) using the three dimensions from `references/evaluation-criteria.md`:
70
71	Scoring Method:
72	1. Quickly assess each dimension (Development Clarity, Testability, Value Clarity) on a 1-5 scale
73	2. Calculate final score: `round((Development Clarity + Testability + Value Clarity) / 3)`
74	3. Use the scoring criteria tables in `references/evaluation-criteria.md` as reference
75
76	Quick Assessment Focus:
77	- Development Clarity: Is action specific? Scope clear? Dependencies clear?
78	- Testability: Can write test cases? Acceptance criteria present? Value verifiable?
79	- Value Clarity: Value clear? Role correct? Maps to requirements?
80
81	\| Score \| Level \| Action \|
82	\|-------\|-------\|--------\|
83	\| 5 \| Excellent \| Keep, no modification \|
84	\| 4 \| Good \| Keep, may have minor suggestions \|
85	\| 3 \| Passing \| Mark for observation, may need minor adjustments \|
86	\| 2 \| Insufficient \| Must correct \|
87	\| 1 \| Severely insufficient \| Must rewrite \|
88
89	Only Stories scoring ≤ 3 enter Phase 2 detailed evaluation.
90
91	### Phase 2: Multi-Perspective Detailed Evaluation
92
93	For Stories needing review, perform detailed evaluation from three perspectives using the Specific Checkpoints and Common Deduction Patterns defined in `references/evaluation-criteria.md`.
94
95	#### 👨‍💻 Developer Perspective
96
97	Reference: `references/evaluation-criteria.md` - Dimension 1: Development Clarity
98
99	Detailed Checkpoints (from evaluation-criteria.md):
100	- [ ] Is action description specific?
101	- 5 points: "Upload JPG/PNG format images, limited to 5MB"
102	- 3 points: "Upload images"
103	- 1 point: "Handle images"
104	- [ ] Does scope have boundaries?
105	- 5 points: "Edit article title and content"
106	- 3 points: "Edit article"
107	- 1 point: "Manage articles"
108	- [ ] Are dependencies clear?
109	- 5 points: Clearly marked "requires US-001 login feature completed first"
110	- 3 points: Implied dependency but not marked
111	- 1 point: Confusing or circular dependencies
112
113	Common Problems (see evaluation-criteria.md for deduction patterns):
114	- Vague verbs: "manage", "handle", "maintain" (-1~2 points)
115	- No scope boundary: "all settings", "various reports" (-1~2 points)
116	- Compound features: "create and edit" (-1 point)
117	- Technical details mixed in: "load using AJAX" (-1 point)
118
119	#### 🧪 QA Perspective
120
121	Reference: `references/evaluation-criteria.md` - Dimension 2: Testability
122
123	Detailed Checkpoints (from evaluation-criteria.md):
124	- [ ] Are acceptance criteria clear?
125	- 5 points: Has specific Given-When-Then or checklist
126	- 3 points: Has general direction but not specific
127	- 1 point: No acceptance criteria, or vague like "should be user-friendly"
128	- [ ] Is value verifiable?
129	- 5 points: "so that I can find target artic

People who liked this also viewed

6 picks

Type