launch-sub-agent

Name: launch-sub-agent
Author: NeoLabHQ

866installs132forks

How to install launch-sub-agent?

neolabhq/context-engineering-kit/launch-sub-agent

$npx -y skills add NeoLabHQ/context-engineering-kit --skill launch-sub-agent

Installs into the current project.

Summary

Launch an intelligent sub-agent with automatic model selection based on task complexity, specialized agent matching, Zero-shot CoT reasoning, and mandatory self-critique verification

Files · 1

View on GitHub

SKILL.md

1	# launch-sub-agent
2
3	<task>
4	Launch a focused sub-agent to execute the provided task. Analyze the task to intelligently select the optimal model and agent configuration, then dispatch a sub-agent with Zero-shot Chain-of-Thought reasoning at the beginning and mandatory self-critique verification at the end.
5	</task>
6
7	<context>
8	This command implements the Supervisor/Orchestrator pattern from multi-agent architectures where you (the orchestrator) dispatch focused sub-agents with isolated context. The primary benefit is context isolation - each sub-agent operates in a clean context window focused on its specific task without accumulated context pollution.
9	</context>
10
11	## Process
12
13	### Phase 1: Task Analysis with Zero-shot CoT
14
15	Before dispatching, analyze the task systematically. Think through step by step:
16
17	```
18	Let me analyze this task step by step to determine the optimal configuration:
19
20	1. Task Type Identification
21	"What type of work is being requested?"
22	- Code implementation / feature development
23	- Research / investigation / comparison
24	- Documentation / technical writing
25	- Code review / quality analysis
26	- Architecture / system design
27	- Testing / validation
28	- Simple transformation / lookup
29
30	2. Complexity Assessment
31	"How complex is the reasoning required?"
32	- High: Architecture decisions, novel problem-solving, multi-faceted analysis
33	- Medium: Standard implementation following patterns, moderate research
34	- Low: Simple transformations, lookups, well-defined single-step tasks
35
36	3. Output Size Estimation
37	"How extensive is the expected output?"
38	- Large: Multiple files, comprehensive documentation, extensive analysis
39	- Medium: Single feature, focused deliverable
40	- Small: Quick answer, minor change, brief output
41
42	4. Domain Expertise Check
43	"Does this task match a specialized agent profile?"
44	- Development: code, implement, feature, endpoint, TDD, tests
45	- Research: investigate, compare, evaluate, options, library
46	- Documentation: document, README, guide, explain, tutorial
47	- Architecture: design, system, structure, scalability
48	- Exploration: understand, navigate, find, codebase patterns
49	```
50
51	### Phase 2: Model Selection
52
53	Select the optimal model based on task analysis:
54
55	\| Task Profile \| Recommended Model \| Rationale \|
56	\|--------------\|-------------------\|-----------\|
57	\| Complex reasoning (architecture, design, critical decisions) \| `opus` \| Maximum reasoning capability \|
58	\| Specialized domain (matches agent profile) \| Opus + Specialized Agent \| Domain expertise + reasoning power \|
59	\| Non-complex but long (extensive docs, verbose output) \| `sonnet[1m]` \| Good capability, cost-efficient for length \|
60	\| Simple and short (trivial tasks, quick lookups) \| `haiku` \| Fast, cost-effective for easy tasks \|
61	\| Default (when uncertain) \| `opus` \| Optimize for quality over cost \|
62
63	Decision Tree:
64
65	```
66	Is task COMPLEX (architecture, design, novel problem, critical decision)?
67	\|
68	+-- YES --> Use Opus (highest capability)
69	\| \|
70	\| +-- Does it match a specialized domain?
71	\| +-- YES --> Include specialized agent prompt
72	\| +-- NO --> Use Opus alone
73	\|
74	+-- NO --> Is task SIMPLE and SHORT?
75	\|
76	+-- YES --> Use Haiku (fast, cheap)
77	\|
78	+-- NO --> Is output LONG but task not complex?
79	\|
80	+-- YES --> Use Sonnet (balanced)
81	\|
82	+-- NO --> Use Opus (default)
83	```
84
85	### Phase 3: Specialized Agent Matching
86
87	If the task matches a specialized domain, incorporate the relevant agent prompt. Specialized agents provide domain-specific best practices, quality standards, and structured approaches that improve output quality.
88
89	Decision: Use specialized agent when task clearly benefits from domain expertise. Skip for trivial tasks where specialization adds unnecessary overhead.
90
91	Agents: Available specialized agents depends on project and plugins installed. Common agents from the `sdd` plugin include: `sdd:developer`, `sdd:researcher`, `sdd:software-architect`, `sdd:tech-lead`, `sdd:team-lead`, `sdd:qa-engineer`, `sdd:code-explorer`, `sdd:business-analyst`. If the appropriate specialized agent is not available, fallback to a general agent without specialization.
92
93	Integration with Model Selection:
94
95	- Specialized agents are combined WITH model selection, not instead of
96	- Complex task + specialized domain = Opus + Specialized Agent
97	- Simple task matching domain = Haiku without specialization (overhead

People who liked this also viewed

6 picks

Type

launch-sub-agent

866installs132forks

View on GitHub

How to install launch-sub-agent?

neolabhq/context-engineering-kit/launch-sub-agent

$npx -y skills add NeoLabHQ/context-engineering-kit --skill launch-sub-agent

Installs into the current project.

Summary

Launch an intelligent sub-agent with automatic model selection based on task complexity, specialized agent matching, Zero-shot CoT reasoning, and mandatory self-critique verification

Files · 1

View on GitHub

SKILL.md

1	# launch-sub-agent
2
3	<task>
4	Launch a focused sub-agent to execute the provided task. Analyze the task to intelligently select the optimal model and agent configuration, then dispatch a sub-agent with Zero-shot Chain-of-Thought reasoning at the beginning and mandatory self-critique verification at the end.
5	</task>
6
7	<context>
8	This command implements the Supervisor/Orchestrator pattern from multi-agent architectures where you (the orchestrator) dispatch focused sub-agents with isolated context. The primary benefit is context isolation - each sub-agent operates in a clean context window focused on its specific task without accumulated context pollution.
9	</context>
10
11	## Process
12
13	### Phase 1: Task Analysis with Zero-shot CoT
14
15	Before dispatching, analyze the task systematically. Think through step by step:
16
17	```
18	Let me analyze this task step by step to determine the optimal configuration:
19
20	1. Task Type Identification
21	"What type of work is being requested?"
22	- Code implementation / feature development
23	- Research / investigation / comparison
24	- Documentation / technical writing
25	- Code review / quality analysis
26	- Architecture / system design
27	- Testing / validation
28	- Simple transformation / lookup
29
30	2. Complexity Assessment
31	"How complex is the reasoning required?"
32	- High: Architecture decisions, novel problem-solving, multi-faceted analysis
33	- Medium: Standard implementation following patterns, moderate research
34	- Low: Simple transformations, lookups, well-defined single-step tasks
35
36	3. Output Size Estimation
37	"How extensive is the expected output?"
38	- Large: Multiple files, comprehensive documentation, extensive analysis
39	- Medium: Single feature, focused deliverable
40	- Small: Quick answer, minor change, brief output
41
42	4. Domain Expertise Check
43	"Does this task match a specialized agent profile?"
44	- Development: code, implement, feature, endpoint, TDD, tests
45	- Research: investigate, compare, evaluate, options, library
46	- Documentation: document, README, guide, explain, tutorial
47	- Architecture: design, system, structure, scalability
48	- Exploration: understand, navigate, find, codebase patterns
49	```
50
51	### Phase 2: Model Selection
52
53	Select the optimal model based on task analysis:
54
55	\| Task Profile \| Recommended Model \| Rationale \|
56	\|--------------\|-------------------\|-----------\|
57	\| Complex reasoning (architecture, design, critical decisions) \| `opus` \| Maximum reasoning capability \|
58	\| Specialized domain (matches agent profile) \| Opus + Specialized Agent \| Domain expertise + reasoning power \|
59	\| Non-complex but long (extensive docs, verbose output) \| `sonnet[1m]` \| Good capability, cost-efficient for length \|
60	\| Simple and short (trivial tasks, quick lookups) \| `haiku` \| Fast, cost-effective for easy tasks \|
61	\| Default (when uncertain) \| `opus` \| Optimize for quality over cost \|
62
63	Decision Tree:
64
65	```
66	Is task COMPLEX (architecture, design, novel problem, critical decision)?
67	\|
68	+-- YES --> Use Opus (highest capability)
69	\| \|
70	\| +-- Does it match a specialized domain?
71	\| +-- YES --> Include specialized agent prompt
72	\| +-- NO --> Use Opus alone
73	\|
74	+-- NO --> Is task SIMPLE and SHORT?
75	\|
76	+-- YES --> Use Haiku (fast, cheap)
77	\|
78	+-- NO --> Is output LONG but task not complex?
79	\|
80	+-- YES --> Use Sonnet (balanced)
81	\|
82	+-- NO --> Use Opus (default)
83	```
84
85	### Phase 3: Specialized Agent Matching
86
87	If the task matches a specialized domain, incorporate the relevant agent prompt. Specialized agents provide domain-specific best practices, quality standards, and structured approaches that improve output quality.
88
89	Decision: Use specialized agent when task clearly benefits from domain expertise. Skip for trivial tasks where specialization adds unnecessary overhead.
90
91	Agents: Available specialized agents depends on project and plugins installed. Common agents from the `sdd` plugin include: `sdd:developer`, `sdd:researcher`, `sdd:software-architect`, `sdd:tech-lead`, `sdd:team-lead`, `sdd:qa-engineer`, `sdd:code-explorer`, `sdd:business-analyst`. If the appropriate specialized agent is not available, fallback to a general agent without specialization.
92
93	Integration with Model Selection:
94
95	- Specialized agents are combined WITH model selection, not instead of
96	- Complex task + specialized domain = Opus + Specialized Agent
97	- Simple task matching domain = Haiku without specialization (overhead

People who liked this also viewed

6 picks

Type