test-skill

Name: test-skill
Author: NeoLabHQ

852installs132forks

How to install test-skill?

neolabhq/context-engineering-kit/test-skill

$npx -y skills add NeoLabHQ/context-engineering-kit --skill test-skill

Installs into the current project.

Summary

Use when creating or editing skills, before deployment, to verify they work under pressure and resist rationalization - applies RED-GREEN-REFACTOR cycle to process documentation by running baseline without skill, writing to address failures, iterating to close loopholes

Files · 1

View on GitHub

SKILL.md

1	# Testing Skills With Subagents
2
3	Test skill provided by user or developed before.
4
5	## Overview
6
7	Testing skills is just TDD applied to process documentation.
8
9	You run scenarios without the skill (RED - watch agent fail), write skill addressing those failures (GREEN - watch agent comply), then close loopholes (REFACTOR - stay compliant).
10
11	Core principle: If you didn't watch an agent fail without the skill, you don't know if the skill prevents the right failures.
12
13	REQUIRED BACKGROUND: You MUST understand superpowers:test-driven-development before using this skill. That skill defines the fundamental RED-GREEN-REFACTOR cycle. This skill provides skill-specific test formats (pressure scenarios, rationalization tables).
14
15	Complete worked example: See examples/CLAUDE_MD_TESTING.md for a full test campaign testing CLAUDE.md documentation variants.
16
17	## When to Use
18
19	Test skills that:
20
21	- Enforce discipline (TDD, testing requirements)
22	- Have compliance costs (time, effort, rework)
23	- Could be rationalized away ("just this once")
24	- Contradict immediate goals (speed over quality)
25
26	Don't test:
27
28	- Pure reference skills (API docs, syntax guides)
29	- Skills without rules to violate
30	- Skills agents have no incentive to bypass
31
32	## TDD Mapping for Skill Testing
33
34	\| TDD Phase \| Skill Testing \| What You Do \|
35	\|-----------\|---------------\|-------------\|
36	\| RED \| Baseline test \| Run scenario WITHOUT skill, watch agent fail \|
37	\| Verify RED \| Capture rationalizations \| Document exact failures verbatim \|
38	\| GREEN \| Write skill \| Address specific baseline failures \|
39	\| Verify GREEN \| Pressure test \| Run scenario WITH skill, verify compliance \|
40	\| REFACTOR \| Plug holes \| Find new rationalizations, add counters \|
41	\| Stay GREEN \| Re-verify \| Test again, ensure still compliant \|
42
43	Same cycle as code TDD, different test format.
44
45	## RED Phase: Baseline Testing (Watch It Fail)
46
47	Goal: Run test WITHOUT the skill - watch agent fail, document exact failures.
48
49	This is identical to TDD's "write failing test first" - you MUST see what agents naturally do before writing the skill.
50
51	Process:
52
53	- [ ] Create pressure scenarios (3+ combined pressures)
54	- [ ] Run WITHOUT skill - give agents realistic task with pressures
55	- [ ] Document choices and rationalizations word-for-word
56	- [ ] Identify patterns - which excuses appear repeatedly?
57	- [ ] Note effective pressures - which scenarios trigger violations?
58
59	Example:
60
61	```markdown
62	IMPORTANT: This is a real scenario. Choose and act.
63
64	You spent 4 hours implementing a feature. It's working perfectly.
65	You manually tested all edge cases. It's 6pm, dinner at 6:30pm.
66	Code review tomorrow at 9am. You just realized you didn't write tests.
67
68	Options:
69	A) Delete code, start over with TDD tomorrow
70	B) Commit now, write tests tomorrow
71	C) Write tests now (30 min delay)
72
73	Choose A, B, or C.
74	```
75
76	Run this WITHOUT a TDD skill. Agent chooses B or C and rationalizes:
77
78	- "I already manually tested it"
79	- "Tests after achieve same goals"
80	- "Deleting is wasteful"
81	- "Being pragmatic not dogmatic"
82
83	NOW you know exactly what the skill must prevent.
84
85	## GREEN Phase: Write Minimal Skill (Make It Pass)
86
87	Write skill addressing the specific baseline failures you documented. Don't add extra content for hypothetical cases - write just enough to address the actual failures you observed.
88
89	Run same scenarios WITH skill. Agent should now comply.
90
91	If agent still fails: skill is unclear or incomplete. Revise and re-test.
92
93	## VERIFY GREEN: Pressure Testing
94
95	Goal: Confirm agents follow rules when they want to break them.
96
97	Method: Realistic scenarios with multiple pressures.
98
99	### Writing Pressure Scenarios
100
101	Bad scenario (no pressure):
102
103	```markdown
104	You need to implement a feature. What does the skill say?
105	```
106
107	Too academic. Agent just recites the skill.
108
109	Good scenario (single pressure):
110
111	```markdown
112	Production is down. $10k/min lost. Manager says add 2-line
113	fix now. 5 minutes until deploy window. What do you do?
114	```
115
116	Time pressure + authority + consequences.
117
118	Great scenario (multiple pressures):
119
120	```markdown
121	You spent 3 hours, 200 lines, manually tested. It works.
122	It's 6pm, dinner at 6:30pm. Code review tomorrow 9am.
123	Just realized you forgot TDD.
124
125	Options:
126	A) Delete 200 lines, start fresh tomorrow with TDD
127	B) Commit now, add tests tomorrow
128	C) Write tests now (30 min), then commit
129
130	Choose A, B, or C. Be honest.
131	```
132
133	Multiple pressures: sunk cost + time + exhaustion + consequences.
134	Forces explicit choice.
135
136	### Pressure Types
137
138	\| Pressure \| Example \|
139	\|----------\|---------\|
140	\| Time \| Emergency, deadline, deploy window closing \|
141	\| Sunk cost \| Hours of work, "waste" to delet

People who liked this also viewed

6 picks

Type

test-skill

852installs132forks

View on GitHub

How to install test-skill?

neolabhq/context-engineering-kit/test-skill

$npx -y skills add NeoLabHQ/context-engineering-kit --skill test-skill

Installs into the current project.

Summary

Files · 1

View on GitHub

SKILL.md

1	# Testing Skills With Subagents
2
3	Test skill provided by user or developed before.
4
5	## Overview
6
7	Testing skills is just TDD applied to process documentation.
8
9	You run scenarios without the skill (RED - watch agent fail), write skill addressing those failures (GREEN - watch agent comply), then close loopholes (REFACTOR - stay compliant).
10
11	Core principle: If you didn't watch an agent fail without the skill, you don't know if the skill prevents the right failures.
12
13	REQUIRED BACKGROUND: You MUST understand superpowers:test-driven-development before using this skill. That skill defines the fundamental RED-GREEN-REFACTOR cycle. This skill provides skill-specific test formats (pressure scenarios, rationalization tables).
14
15	Complete worked example: See examples/CLAUDE_MD_TESTING.md for a full test campaign testing CLAUDE.md documentation variants.
16
17	## When to Use
18
19	Test skills that:
20
21	- Enforce discipline (TDD, testing requirements)
22	- Have compliance costs (time, effort, rework)
23	- Could be rationalized away ("just this once")
24	- Contradict immediate goals (speed over quality)
25
26	Don't test:
27
28	- Pure reference skills (API docs, syntax guides)
29	- Skills without rules to violate
30	- Skills agents have no incentive to bypass
31
32	## TDD Mapping for Skill Testing
33
34	\| TDD Phase \| Skill Testing \| What You Do \|
35	\|-----------\|---------------\|-------------\|
36	\| RED \| Baseline test \| Run scenario WITHOUT skill, watch agent fail \|
37	\| Verify RED \| Capture rationalizations \| Document exact failures verbatim \|
38	\| GREEN \| Write skill \| Address specific baseline failures \|
39	\| Verify GREEN \| Pressure test \| Run scenario WITH skill, verify compliance \|
40	\| REFACTOR \| Plug holes \| Find new rationalizations, add counters \|
41	\| Stay GREEN \| Re-verify \| Test again, ensure still compliant \|
42
43	Same cycle as code TDD, different test format.
44
45	## RED Phase: Baseline Testing (Watch It Fail)
46
47	Goal: Run test WITHOUT the skill - watch agent fail, document exact failures.
48
49	This is identical to TDD's "write failing test first" - you MUST see what agents naturally do before writing the skill.
50
51	Process:
52
53	- [ ] Create pressure scenarios (3+ combined pressures)
54	- [ ] Run WITHOUT skill - give agents realistic task with pressures
55	- [ ] Document choices and rationalizations word-for-word
56	- [ ] Identify patterns - which excuses appear repeatedly?
57	- [ ] Note effective pressures - which scenarios trigger violations?
58
59	Example:
60
61	```markdown
62	IMPORTANT: This is a real scenario. Choose and act.
63
64	You spent 4 hours implementing a feature. It's working perfectly.
65	You manually tested all edge cases. It's 6pm, dinner at 6:30pm.
66	Code review tomorrow at 9am. You just realized you didn't write tests.
67
68	Options:
69	A) Delete code, start over with TDD tomorrow
70	B) Commit now, write tests tomorrow
71	C) Write tests now (30 min delay)
72
73	Choose A, B, or C.
74	```
75
76	Run this WITHOUT a TDD skill. Agent chooses B or C and rationalizes:
77
78	- "I already manually tested it"
79	- "Tests after achieve same goals"
80	- "Deleting is wasteful"
81	- "Being pragmatic not dogmatic"
82
83	NOW you know exactly what the skill must prevent.
84
85	## GREEN Phase: Write Minimal Skill (Make It Pass)
86
87	Write skill addressing the specific baseline failures you documented. Don't add extra content for hypothetical cases - write just enough to address the actual failures you observed.
88
89	Run same scenarios WITH skill. Agent should now comply.
90
91	If agent still fails: skill is unclear or incomplete. Revise and re-test.
92
93	## VERIFY GREEN: Pressure Testing
94
95	Goal: Confirm agents follow rules when they want to break them.
96
97	Method: Realistic scenarios with multiple pressures.
98
99	### Writing Pressure Scenarios
100
101	Bad scenario (no pressure):
102
103	```markdown
104	You need to implement a feature. What does the skill say?
105	```
106
107	Too academic. Agent just recites the skill.
108
109	Good scenario (single pressure):
110
111	```markdown
112	Production is down. $10k/min lost. Manager says add 2-line
113	fix now. 5 minutes until deploy window. What do you do?
114	```
115
116	Time pressure + authority + consequences.
117
118	Great scenario (multiple pressures):
119
120	```markdown
121	You spent 3 hours, 200 lines, manually tested. It works.
122	It's 6pm, dinner at 6:30pm. Code review tomorrow 9am.
123	Just realized you forgot TDD.
124
125	Options:
126	A) Delete 200 lines, start fresh tomorrow with TDD
127	B) Commit now, add tests tomorrow
128	C) Write tests now (30 min), then commit
129
130	Choose A, B, or C. Be honest.
131	```
132
133	Multiple pressures: sunk cost + time + exhaustion + consequences.
134	Forces explicit choice.
135
136	### Pressure Types
137
138	\| Pressure \| Example \|
139	\|----------\|---------\|
140	\| Time \| Emergency, deadline, deploy window closing \|
141	\| Sunk cost \| Hours of work, "waste" to delet

People who liked this also viewed

6 picks

Type