agent-selector

Name: agent-selector
Author: DNYoussef

32stars6forks

How to install agent-selector?

dnyoussef/context-cascade/agent-selector

$npx -y skills add DNYoussef/context-cascade --skill agent-selector

Installs into the current project.

Summary

/*============================================================================*/ /* AGENT-SELECTOR SKILL :: VERILINGUA x VERIX EDITION */ /*============================================================================*/

Files · 1

View on GitHub

SKILL.md

1	/============================================================================/
2	/* AGENT-SELECTOR SKILL :: VERILINGUA x VERIX EDITION */
3	/============================================================================/
4
5	---
6	name: agent-selector
7	version: 2.1.0
8	description: \|
9	[assert\|neutral] Intelligent agent selection from 203-agent registry using semantic matching and capability analysis [ground:given] [conf:0.95] [state:confirmed]
10	category: orchestration
11	tags:
12	- general
13	author: System
14	cognitive_frame:
15	primary: evidential
16	goal_analysis:
17	first_order: "Execute agent-selector workflow"
18	second_order: "Ensure quality and consistency"
19	third_order: "Enable systematic orchestration processes"
20	---
21
22	/----------------------------------------------------------------------------/
23	/* S0 META-IDENTITY */
24	/----------------------------------------------------------------------------/
25
26	[define\|neutral] SKILL := {
27	name: "agent-selector",
28	category: "orchestration",
29	version: "2.1.0",
30	layer: L1
31	} [ground:given] [conf:1.0] [state:confirmed]
32
33	/----------------------------------------------------------------------------/
34	/* S1 COGNITIVE FRAME */
35	/----------------------------------------------------------------------------/
36
37	[define\|neutral] COGNITIVE_FRAME := {
38	frame: "Evidential",
39	source: "Turkish",
40	force: "How do you know?"
41	} [ground:cognitive-science] [conf:0.92] [state:confirmed]
42
43	## Kanitsal Cerceve (Evidential Frame Activation)
44	Kaynak dogrulama modu etkin.
45
46	/----------------------------------------------------------------------------/
47	/* S2 TRIGGER CONDITIONS */
48	/----------------------------------------------------------------------------/
49
50	[define\|neutral] TRIGGER_POSITIVE := {
51	keywords: ["agent-selector", "orchestration", "workflow"],
52	context: "user needs agent-selector capability"
53	} [ground:given] [conf:1.0] [state:confirmed]
54
55	/----------------------------------------------------------------------------/
56	/* S3 CORE CONTENT */
57	/----------------------------------------------------------------------------/
58
59	# Agent Selector Micro-Skill
60
61	## Kanitsal Cerceve (Evidential Frame Activation)
62	Kaynak dogrulama modu etkin.
63
64
65
66	## Phase 0: Expertise Loading
67
68	Before selecting agents:
69
70	1. Detect Domain: Identify task domain from request
71	2. Check Expertise: Look for `.claude/expertise/agent-selection.yaml`
72	3. Load Context: If exists, load agent performance history and preferences
73	4. Apply Configuration: Use expertise for optimal agent matching
74
75	## Purpose
76
77	Intelligently selects the most appropriate specialized agent from the 203-agent registry based on:
78	- Task requirements and complexity
79	- Agent capabilities and specializations
80	- Domain expertise (category/subcategory)
81	- Tool and MCP requirements
82	- Phase alignment (planning, development, testing, etc.)
83
84	Critical for Phase 4 routing to ensure Claude Code uses specialized agents instead of generic ones.
85
86	## When to Use
87
88	- Before any Task() invocation in Phase 5 execution
89	- When planning multi-agent workflows and need optimal agent assignment
90	- When you're unsure which specialized agent to use for a task
91	- To validate that a generic agent name has a specialized alternative
92
93	## How It Works
94
95	4-Step Process:
96
97	1. Parse Task Requirements
98	- Extract domain (backend, frontend, database, testing, etc.)
99	- Identify key capabilities needed (Express.js, PostgreSQL, TDD, etc.)
100	- Determine phase (planning, development, testing, deployment)
101	- Note tool/MCP requirements
102
103	2. Semantic Search (Memory MCP)
104	- Query Memory MCP with task description
105	- Get top 5-10 candidate agents ranked by similarity
106	- Filter by category/phase if known
107
108	3. Capability Matching
109	- Score each candidate agent based on:
110	- Exact capability matches (highest priority)
111	- Domain specialization (category/subcategory)
112	- Tool/MCP alignment
113	- Phase alignment
114	- Apply fallback rules if no perfect match
115
116	4. Return Selection + Reasoning
117	- Selected agent name
118	- Agent source (file path in registry)
119	- Capabilities that matched
120	- Alternatives considered
121	- Selection reasoning
122
123	## Usage
124
125	```javascript
126	// Skill invocation
127	Skill("agent-selector")
128
129	// Agent will prompt you for:
130	// 1. Task description (what needs to be done)
131	// 2. Domain hint (optional: backend, frontend, testing, etc.)
132	// 3. Phase hint (optional: development, testing, deployment)
133
134	// Output:
135	{
136	"selected_agent": "dev-backend-api",
137	"agent_source": "delivery/development/backend/dev-backend-api.md",
138	"agent_category": "delivery/development/backend",
139	"capabilities": ["Express.js", "REST APIs", "JWT", "OpenAPI"],
140	"selection_reasoning": "Specialized backend API agent with exact match for

People who liked this also viewed

6 picks

Type

agent-selector

32stars6forks

View on GitHub

How to install agent-selector?

dnyoussef/context-cascade/agent-selector

$npx -y skills add DNYoussef/context-cascade --skill agent-selector

Installs into the current project.

Summary

Files · 1

View on GitHub

SKILL.md

1	/============================================================================/
2	/* AGENT-SELECTOR SKILL :: VERILINGUA x VERIX EDITION */
3	/============================================================================/
4
5	---
6	name: agent-selector
7	version: 2.1.0
8	description: \|
9	[assert\|neutral] Intelligent agent selection from 203-agent registry using semantic matching and capability analysis [ground:given] [conf:0.95] [state:confirmed]
10	category: orchestration
11	tags:
12	- general
13	author: System
14	cognitive_frame:
15	primary: evidential
16	goal_analysis:
17	first_order: "Execute agent-selector workflow"
18	second_order: "Ensure quality and consistency"
19	third_order: "Enable systematic orchestration processes"
20	---
21
22	/----------------------------------------------------------------------------/
23	/* S0 META-IDENTITY */
24	/----------------------------------------------------------------------------/
25
26	[define\|neutral] SKILL := {
27	name: "agent-selector",
28	category: "orchestration",
29	version: "2.1.0",
30	layer: L1
31	} [ground:given] [conf:1.0] [state:confirmed]
32
33	/----------------------------------------------------------------------------/
34	/* S1 COGNITIVE FRAME */
35	/----------------------------------------------------------------------------/
36
37	[define\|neutral] COGNITIVE_FRAME := {
38	frame: "Evidential",
39	source: "Turkish",
40	force: "How do you know?"
41	} [ground:cognitive-science] [conf:0.92] [state:confirmed]
42
43	## Kanitsal Cerceve (Evidential Frame Activation)
44	Kaynak dogrulama modu etkin.
45
46	/----------------------------------------------------------------------------/
47	/* S2 TRIGGER CONDITIONS */
48	/----------------------------------------------------------------------------/
49
50	[define\|neutral] TRIGGER_POSITIVE := {
51	keywords: ["agent-selector", "orchestration", "workflow"],
52	context: "user needs agent-selector capability"
53	} [ground:given] [conf:1.0] [state:confirmed]
54
55	/----------------------------------------------------------------------------/
56	/* S3 CORE CONTENT */
57	/----------------------------------------------------------------------------/
58
59	# Agent Selector Micro-Skill
60
61	## Kanitsal Cerceve (Evidential Frame Activation)
62	Kaynak dogrulama modu etkin.
63
64
65
66	## Phase 0: Expertise Loading
67
68	Before selecting agents:
69
70	1. Detect Domain: Identify task domain from request
71	2. Check Expertise: Look for `.claude/expertise/agent-selection.yaml`
72	3. Load Context: If exists, load agent performance history and preferences
73	4. Apply Configuration: Use expertise for optimal agent matching
74
75	## Purpose
76
77	Intelligently selects the most appropriate specialized agent from the 203-agent registry based on:
78	- Task requirements and complexity
79	- Agent capabilities and specializations
80	- Domain expertise (category/subcategory)
81	- Tool and MCP requirements
82	- Phase alignment (planning, development, testing, etc.)
83
84	Critical for Phase 4 routing to ensure Claude Code uses specialized agents instead of generic ones.
85
86	## When to Use
87
88	- Before any Task() invocation in Phase 5 execution
89	- When planning multi-agent workflows and need optimal agent assignment
90	- When you're unsure which specialized agent to use for a task
91	- To validate that a generic agent name has a specialized alternative
92
93	## How It Works
94
95	4-Step Process:
96
97	1. Parse Task Requirements
98	- Extract domain (backend, frontend, database, testing, etc.)
99	- Identify key capabilities needed (Express.js, PostgreSQL, TDD, etc.)
100	- Determine phase (planning, development, testing, deployment)
101	- Note tool/MCP requirements
102
103	2. Semantic Search (Memory MCP)
104	- Query Memory MCP with task description
105	- Get top 5-10 candidate agents ranked by similarity
106	- Filter by category/phase if known
107
108	3. Capability Matching
109	- Score each candidate agent based on:
110	- Exact capability matches (highest priority)
111	- Domain specialization (category/subcategory)
112	- Tool/MCP alignment
113	- Phase alignment
114	- Apply fallback rules if no perfect match
115
116	4. Return Selection + Reasoning
117	- Selected agent name
118	- Agent source (file path in registry)
119	- Capabilities that matched
120	- Alternatives considered
121	- Selection reasoning
122
123	## Usage
124
125	```javascript
126	// Skill invocation
127	Skill("agent-selector")
128
129	// Agent will prompt you for:
130	// 1. Task description (what needs to be done)
131	// 2. Domain hint (optional: backend, frontend, testing, etc.)
132	// 3. Phase hint (optional: development, testing, deployment)
133
134	// Output:
135	{
136	"selected_agent": "dev-backend-api",
137	"agent_source": "delivery/development/backend/dev-backend-api.md",
138	"agent_category": "delivery/development/backend",
139	"capabilities": ["Express.js", "REST APIs", "JWT", "OpenAPI"],
140	"selection_reasoning": "Specialized backend API agent with exact match for

People who liked this also viewed

6 picks

Type