elicitation

Name: elicitation
Author: fusengine

21stars3forks

How to install elicitation?

fusengine/agents/elicitation

$npx -y skills add fusengine/agents --skill elicitation

Installs into the current project.

Summary

Auto-review skill for expert agents. After coding, expert applies elicitation techniques to self-correct before sniper validation. Inspired by BMAD-METHOD. Use when: an expert agent needs to self-review and self-correct code after the Execute phase, before sniper validation.

Files · 1

View on GitHub

SKILL.md

1	# Elicitation Skill
2
3	## Purpose
4
5	Enable expert agents to self-review and self-correct their code before external validation (sniper). Based on BMAD-METHOD's 75 elicitation techniques.
6
7	---
8
9	## 3 Execution Modes
10
11	### Mode 1: MANUAL (default)
12	```
13	Expert presents 5 relevant techniques → User chooses → Expert applies
14	```
15
16	### Mode 2: AUTO (--auto)
17	```
18	Expert auto-detects code type → Auto-selects techniques → Applies silently
19	```
20
21	### Mode 3: SKIP (--skip)
22	```
23	Skip elicitation → Go directly to sniper validation
24	```
25
26	---
27
28	## Quick Start
29
30	After Execute phase, expert runs:
31	```bash
32	# Manual mode (default)
33	> Apply elicitation skill
34
35	# Auto mode (no prompts)
36	> Apply elicitation skill --auto
37
38	# Skip self-review
39	> Apply elicitation skill --skip
40	```
41
42	---
43
44	## Workflow Overview
45
46	```
47	┌─────────────────────────────────────────────────────────┐
48	│ ELICITATION WORKFLOW │
49	│ │
50	│ Step 0: Init → Load context │
51	│ Step 1: Analyze Code → Detect code type │
52	│ Step 2: Select → Choose techniques (or auto) │
53	│ Step 3: Apply Review → Execute techniques │
54	│ Step 4: Self-Correct → Fix own issues │
55	│ Step 5: Report → Summary before sniper │
56	└─────────────────────────────────────────────────────────┘
57	```
58
59	---
60
61	## Auto-Detection Matrix
62
63	\| Code Type Detected \| Auto-Selected Techniques \|
64	\|--------------------\|--------------------------\|
65	\| Authentication/Security \| Security Audit, OWASP Check, Input Validation \|
66	\| API Endpoints \| Error Handling, Type Coverage, API Contracts \|
67	\| Database/ORM \| N+1 Detection, Migration Safety, Data Integrity \|
68	\| UI Components \| Accessibility, Edge Cases, Loading States \|
69	\| Business Logic \| SOLID Compliance, Unit Test Coverage, Edge Cases \|
70	\| Refactoring \| Breaking Changes, Regression Analysis, Backward Compat \|
71	\| Performance Critical \| Profiling, Memory Analysis, Complexity Check \|
72	\| Config/Docs/Plugin files (`.md` agents/skills, `hooks.json`, frontmatter YAML) \| CQ-01, DOC-01, INT-01 + validation by the strictest parser in the consumption chain (e.g. `js-yaml` strict for frontmatter, `json.tool` for JSON) -- never "looks well-formed" \|
73
74	---
75
76	## Technique Categories (12)
77
78	Full catalog: `references/techniques-catalog.md`
79
80	1. Code Quality (7): Code review, Pattern detection, Complexity analysis...
81	2. Security (7): OWASP audit, Input validation, Auth check...
82	3. Performance (6): Profiling, N+1 detection, Memory analysis...
83	4. Architecture (6): SOLID check, Dependency analysis, Coupling review...
84	5. Testing (6): Edge cases, Boundary testing, Error paths...
85	6. Documentation (6): API review, Comment check, Type coverage...
86	7. UX (6): Accessibility, Error messages, Loading states...
87	8. Data (6): Schema validation, Migration safety, Data integrity...
88	9. Concurrency (6): Race conditions, Deadlock analysis, State sync...
89	10. Integration (7): API contracts, Backward compat, Breaking changes...
90	11. Observability (6): Logging, Metrics, Error tracking...
91	12. Maintainability (6): Readability, Naming, File organization...
92
93	Total: 75 techniques
94
95	---
96
97	## Integration with APEX
98
99	```
100	A-nalyze → P-lan → E-xecute → [ELICIT] → X-amine
101	│ │ │
102	▼ ▼ ▼
103	Expert Expert sniper
104	code self-review (final)
105	```
106
107	Benefits:
108	- Expert catches own mistakes before sniper
109	- Faster validation (less sniper corrections)
110	- Knowledge retention (expert learns from self-review)
111
112	---
113
114	## Forbidden
115
116	- ❌ Skip init step (must load context)
117	- ❌ Apply techniques without understanding code type
118	- ❌ Self-correct without documenting changes
119	- ❌ Report without listing applied techniques
120	- ❌ Use techniques outside expertise domain
121
122	---
123
124	## Exit Criteria (Step 5: Report)
125
126	Score = techniques applied / techniques selected, weighted by critical category (Security, Architecture).
127
128	\| Score \| Status \| Action \|
129	\|-------\|--------\|--------\|
130	\| ≥ 90% \| 🟢 \| Proceed to sniper \|
131	\| 70-89% \| 🟡 \| Document gaps, then proceed \|
132	\| < 70% \| 🔴 \| Iterate before sniper \|
133
134	Self-correction failure: If a self-correction breaks the code → revert that correction and keep the finding as a report item instead.
135
136	---
137
138	## Artifact Contract
139
140	Step 5 persists `.claude/apex/docs/elicit-{task-slug}.json` so a later pass
141	diffs against prior verdicts instead of restarting from scratch. Full
142	contract, `{task-slug}` derivation, and

People who liked this also viewed

6 picks

Type

elicitation

21stars3forks

View on GitHub

How to install elicitation?

fusengine/agents/elicitation

$npx -y skills add fusengine/agents --skill elicitation

Installs into the current project.

Summary

Files · 1

View on GitHub

SKILL.md

1	# Elicitation Skill
2
3	## Purpose
4
5	Enable expert agents to self-review and self-correct their code before external validation (sniper). Based on BMAD-METHOD's 75 elicitation techniques.
6
7	---
8
9	## 3 Execution Modes
10
11	### Mode 1: MANUAL (default)
12	```
13	Expert presents 5 relevant techniques → User chooses → Expert applies
14	```
15
16	### Mode 2: AUTO (--auto)
17	```
18	Expert auto-detects code type → Auto-selects techniques → Applies silently
19	```
20
21	### Mode 3: SKIP (--skip)
22	```
23	Skip elicitation → Go directly to sniper validation
24	```
25
26	---
27
28	## Quick Start
29
30	After Execute phase, expert runs:
31	```bash
32	# Manual mode (default)
33	> Apply elicitation skill
34
35	# Auto mode (no prompts)
36	> Apply elicitation skill --auto
37
38	# Skip self-review
39	> Apply elicitation skill --skip
40	```
41
42	---
43
44	## Workflow Overview
45
46	```
47	┌─────────────────────────────────────────────────────────┐
48	│ ELICITATION WORKFLOW │
49	│ │
50	│ Step 0: Init → Load context │
51	│ Step 1: Analyze Code → Detect code type │
52	│ Step 2: Select → Choose techniques (or auto) │
53	│ Step 3: Apply Review → Execute techniques │
54	│ Step 4: Self-Correct → Fix own issues │
55	│ Step 5: Report → Summary before sniper │
56	└─────────────────────────────────────────────────────────┘
57	```
58
59	---
60
61	## Auto-Detection Matrix
62
63	\| Code Type Detected \| Auto-Selected Techniques \|
64	\|--------------------\|--------------------------\|
65	\| Authentication/Security \| Security Audit, OWASP Check, Input Validation \|
66	\| API Endpoints \| Error Handling, Type Coverage, API Contracts \|
67	\| Database/ORM \| N+1 Detection, Migration Safety, Data Integrity \|
68	\| UI Components \| Accessibility, Edge Cases, Loading States \|
69	\| Business Logic \| SOLID Compliance, Unit Test Coverage, Edge Cases \|
70	\| Refactoring \| Breaking Changes, Regression Analysis, Backward Compat \|
71	\| Performance Critical \| Profiling, Memory Analysis, Complexity Check \|
72	\| Config/Docs/Plugin files (`.md` agents/skills, `hooks.json`, frontmatter YAML) \| CQ-01, DOC-01, INT-01 + validation by the strictest parser in the consumption chain (e.g. `js-yaml` strict for frontmatter, `json.tool` for JSON) -- never "looks well-formed" \|
73
74	---
75
76	## Technique Categories (12)
77
78	Full catalog: `references/techniques-catalog.md`
79
80	1. Code Quality (7): Code review, Pattern detection, Complexity analysis...
81	2. Security (7): OWASP audit, Input validation, Auth check...
82	3. Performance (6): Profiling, N+1 detection, Memory analysis...
83	4. Architecture (6): SOLID check, Dependency analysis, Coupling review...
84	5. Testing (6): Edge cases, Boundary testing, Error paths...
85	6. Documentation (6): API review, Comment check, Type coverage...
86	7. UX (6): Accessibility, Error messages, Loading states...
87	8. Data (6): Schema validation, Migration safety, Data integrity...
88	9. Concurrency (6): Race conditions, Deadlock analysis, State sync...
89	10. Integration (7): API contracts, Backward compat, Breaking changes...
90	11. Observability (6): Logging, Metrics, Error tracking...
91	12. Maintainability (6): Readability, Naming, File organization...
92
93	Total: 75 techniques
94
95	---
96
97	## Integration with APEX
98
99	```
100	A-nalyze → P-lan → E-xecute → [ELICIT] → X-amine
101	│ │ │
102	▼ ▼ ▼
103	Expert Expert sniper
104	code self-review (final)
105	```
106
107	Benefits:
108	- Expert catches own mistakes before sniper
109	- Faster validation (less sniper corrections)
110	- Knowledge retention (expert learns from self-review)
111
112	---
113
114	## Forbidden
115
116	- ❌ Skip init step (must load context)
117	- ❌ Apply techniques without understanding code type
118	- ❌ Self-correct without documenting changes
119	- ❌ Report without listing applied techniques
120	- ❌ Use techniques outside expertise domain
121
122	---
123
124	## Exit Criteria (Step 5: Report)
125
126	Score = techniques applied / techniques selected, weighted by critical category (Security, Architecture).
127
128	\| Score \| Status \| Action \|
129	\|-------\|--------\|--------\|
130	\| ≥ 90% \| 🟢 \| Proceed to sniper \|
131	\| 70-89% \| 🟡 \| Document gaps, then proceed \|
132	\| < 70% \| 🔴 \| Iterate before sniper \|
133
134	Self-correction failure: If a self-correction breaks the code → revert that correction and keep the finding as a report item instead.
135
136	---
137
138	## Artifact Contract
139
140	Step 5 persists `.claude/apex/docs/elicit-{task-slug}.json` so a later pass
141	diffs against prior verdicts instead of restarting from scratch. Full
142	contract, `{task-slug}` derivation, and

People who liked this also viewed

6 picks

Type