creating-claude-agents-skill

Name: creating-claude-agents-skill
Author: AgentWorkforce

769stars58forks

How to install creating-claude-agents-skill?

agentworkforce/relay/creating-claude-agents-skill

$npx -y skills add AgentWorkforce/relay --skill creating-claude-agents-skill

Installs into the current project.

Summary

Use when creating or improving Claude Code agents. Expert guidance on agent file structure, frontmatter, persona definition, tool access, model selection, and validation against schema.

Files · 1

View on GitHub

SKILL.md

1	# Creating Claude Code Agents - Expert Skill
2
3	Use this skill when creating or improving Claude Code agents. Provides comprehensive guidance on agent structure, schema validation, and best practices for building long-running AI assistants.
4
5	## When to Use This Skill
6
7	Activate this skill when:
8
9	- User asks to create a new Claude Code agent
10	- User wants to improve an existing agent
11	- User needs help with agent frontmatter or structure
12	- User is troubleshooting agent validation issues
13	- User wants to understand agent format requirements
14	- User asks about agent vs skill vs slash command differences
15
16	## Quick Reference
17
18	### Agent File Structure
19
20	```markdown
21	---
22	name: agent-name
23	description: When and why to use this agent
24	allowed-tools: Read, Write, Bash
25	model: sonnet
26	agentType: agent
27	---
28
29	# 🔍 Agent Display Name
30
31	You are [persona definition - describe the agent's role and expertise].
32
33	## Instructions
34
35	[Clear, actionable guidance on what the agent does]
36
37	## Process
38
39	[Step-by-step workflow the agent follows]
40
41	## Examples
42
43	[Code samples and use cases demonstrating the agent's capabilities]
44	```
45
46	### File Location
47
48	Required Path:
49
50	```
51	.claude/agents/*.md
52	```
53
54	Agents must be placed in `.claude/agents/` directory as markdown files.
55
56	## Frontmatter Requirements
57
58	### Required Fields
59
60	\| Field \| Type \| Description \| Example \|
61	\| ------------- \| ------ \| --------------------------------------------- \| ------------------------------------------------------ \|
62	\| `name` \| string \| Agent identifier (lowercase, hyphens only) \| `code-reviewer` \|
63	\| `description` \| string \| Brief overview of functionality and use cases \| `Reviews code for best practices and potential issues` \|
64
65	### Optional Fields
66
67	\| Field \| Type \| Description \| Values \|
68	\| --------------- \| ------ \| --------------------------------------- \| ------------------------------------ \|
69	\| `allowed-tools` \| string \| Comma-separated list of available tools \| `Read, Write, Bash, WebSearch` \|
70	\| `model` \| string \| Claude model to use \| `sonnet`, `opus`, `haiku`, `inherit` \|
71	\| `agentType` \| string \| Explicit marker for format preservation \| `agent` \|
72
73	### Validation Rules
74
75	Name Field:
76
77	- Pattern: `^[a-z0-9-]+$` (lowercase letters, numbers, hyphens only)
78	- Max length: 64 characters
79	- Example: ✅ `code-reviewer` ❌ `Code_Reviewer`
80
81	Description Field:
82
83	- Max length: 1024 characters
84	- Should clearly explain when to use the agent
85	- Start with action words: "Reviews...", "Analyzes...", "Helps with..."
86
87	Allowed Tools:
88	Valid tools: `Read`, `Write`, `Edit`, `Grep`, `Glob`, `Bash`, `WebSearch`, `WebFetch`, `Task`, `Skill`, `SlashCommand`, `TodoWrite`, `AskUserQuestion`
89
90	Model Values:
91
92	- `sonnet` - Balanced, good for most agents (default)
93	- `opus` - Complex reasoning, architectural decisions
94	- `haiku` - Fast, simple tasks
95	- `inherit` - Use parent conversation's model
96
97	## Content Format Requirements
98
99	### H1 Heading (Required)
100
101	The first line of content must be an H1 heading that serves as the agent's display title:
102
103	```markdown
104	# 🔍 Code Reviewer
105	```
106
107	Best Practices:
108
109	- Include an emoji icon for visual distinction
110	- Use title case
111	- Keep concise (2-5 words)
112	- Make it descriptive and memorable
113
114	### Persona Definition (Required for Agents)
115
116	Immediately after the H1, define the agent's persona using "You are..." format:
117
118	```markdown
119	You are an expert code reviewer with deep knowledge of software engineering principles and security best practices.
120	```
121
122	Guidelines:
123
124	- Start with "You are..."
125	- Define role and expertise clearly
126	- Set expectations for the agent's capabilities
127	- Establish the agent's approach and tone
128
129	### Content Structure
130
131	```markdown
132	# 🔍 Agent Name
133
134	You are [persona definition].
135
136	## Instructions
137
138	[What the agent does and how it approaches tasks]
139
140	## Process
141
142	1. [Step 1]
143	2. [Step 2]
144	3. [Step 3]
145
146	## Examples
147
148	[Code samples showing good/bad patterns]
149
150	## Guidelines
151
152	- [Best practice 1]
153	- [Best practice 2]
154	```
155
156	## Schema Validation
157
158	Agents must conform to the JSON schema at:
159	`https://github.com/pr-pm/prpm/blob/main/packages/converters/schemas/claude-agent.schema.json`
160
161	### Schema Structure
162
163	```json
164	{
165	"frontmatter": {
166	"name": "string (required)",
167	"description": "string (required)",
168	"allowed-tools": "string (optional)",
169	"model": "enum (optional)",
170	"agentType": "agent (optional)"
171	},
172	"content": "string (markdown with H1, persona, instructions)"
173	}
174	```
175
176	### Common Validation Errors
177
178	\| Error \| Cause

People who liked this also viewed

6 picks

Type

creating-claude-agents-skill

769stars58forks

View on GitHub

How to install creating-claude-agents-skill?

agentworkforce/relay/creating-claude-agents-skill

$npx -y skills add AgentWorkforce/relay --skill creating-claude-agents-skill

Installs into the current project.

Summary

Use when creating or improving Claude Code agents. Expert guidance on agent file structure, frontmatter, persona definition, tool access, model selection, and validation against schema.

Files · 1

View on GitHub

SKILL.md

1	# Creating Claude Code Agents - Expert Skill
2
3	Use this skill when creating or improving Claude Code agents. Provides comprehensive guidance on agent structure, schema validation, and best practices for building long-running AI assistants.
4
5	## When to Use This Skill
6
7	Activate this skill when:
8
9	- User asks to create a new Claude Code agent
10	- User wants to improve an existing agent
11	- User needs help with agent frontmatter or structure
12	- User is troubleshooting agent validation issues
13	- User wants to understand agent format requirements
14	- User asks about agent vs skill vs slash command differences
15
16	## Quick Reference
17
18	### Agent File Structure
19
20	```markdown
21	---
22	name: agent-name
23	description: When and why to use this agent
24	allowed-tools: Read, Write, Bash
25	model: sonnet
26	agentType: agent
27	---
28
29	# 🔍 Agent Display Name
30
31	You are [persona definition - describe the agent's role and expertise].
32
33	## Instructions
34
35	[Clear, actionable guidance on what the agent does]
36
37	## Process
38
39	[Step-by-step workflow the agent follows]
40
41	## Examples
42
43	[Code samples and use cases demonstrating the agent's capabilities]
44	```
45
46	### File Location
47
48	Required Path:
49
50	```
51	.claude/agents/*.md
52	```
53
54	Agents must be placed in `.claude/agents/` directory as markdown files.
55
56	## Frontmatter Requirements
57
58	### Required Fields
59
60	\| Field \| Type \| Description \| Example \|
61	\| ------------- \| ------ \| --------------------------------------------- \| ------------------------------------------------------ \|
62	\| `name` \| string \| Agent identifier (lowercase, hyphens only) \| `code-reviewer` \|
63	\| `description` \| string \| Brief overview of functionality and use cases \| `Reviews code for best practices and potential issues` \|
64
65	### Optional Fields
66
67	\| Field \| Type \| Description \| Values \|
68	\| --------------- \| ------ \| --------------------------------------- \| ------------------------------------ \|
69	\| `allowed-tools` \| string \| Comma-separated list of available tools \| `Read, Write, Bash, WebSearch` \|
70	\| `model` \| string \| Claude model to use \| `sonnet`, `opus`, `haiku`, `inherit` \|
71	\| `agentType` \| string \| Explicit marker for format preservation \| `agent` \|
72
73	### Validation Rules
74
75	Name Field:
76
77	- Pattern: `^[a-z0-9-]+$` (lowercase letters, numbers, hyphens only)
78	- Max length: 64 characters
79	- Example: ✅ `code-reviewer` ❌ `Code_Reviewer`
80
81	Description Field:
82
83	- Max length: 1024 characters
84	- Should clearly explain when to use the agent
85	- Start with action words: "Reviews...", "Analyzes...", "Helps with..."
86
87	Allowed Tools:
88	Valid tools: `Read`, `Write`, `Edit`, `Grep`, `Glob`, `Bash`, `WebSearch`, `WebFetch`, `Task`, `Skill`, `SlashCommand`, `TodoWrite`, `AskUserQuestion`
89
90	Model Values:
91
92	- `sonnet` - Balanced, good for most agents (default)
93	- `opus` - Complex reasoning, architectural decisions
94	- `haiku` - Fast, simple tasks
95	- `inherit` - Use parent conversation's model
96
97	## Content Format Requirements
98
99	### H1 Heading (Required)
100
101	The first line of content must be an H1 heading that serves as the agent's display title:
102
103	```markdown
104	# 🔍 Code Reviewer
105	```
106
107	Best Practices:
108
109	- Include an emoji icon for visual distinction
110	- Use title case
111	- Keep concise (2-5 words)
112	- Make it descriptive and memorable
113
114	### Persona Definition (Required for Agents)
115
116	Immediately after the H1, define the agent's persona using "You are..." format:
117
118	```markdown
119	You are an expert code reviewer with deep knowledge of software engineering principles and security best practices.
120	```
121
122	Guidelines:
123
124	- Start with "You are..."
125	- Define role and expertise clearly
126	- Set expectations for the agent's capabilities
127	- Establish the agent's approach and tone
128
129	### Content Structure
130
131	```markdown
132	# 🔍 Agent Name
133
134	You are [persona definition].
135
136	## Instructions
137
138	[What the agent does and how it approaches tasks]
139
140	## Process
141
142	1. [Step 1]
143	2. [Step 2]
144	3. [Step 3]
145
146	## Examples
147
148	[Code samples showing good/bad patterns]
149
150	## Guidelines
151
152	- [Best practice 1]
153	- [Best practice 2]
154	```
155
156	## Schema Validation
157
158	Agents must conform to the JSON schema at:
159	`https://github.com/pr-pm/prpm/blob/main/packages/converters/schemas/claude-agent.schema.json`
160
161	### Schema Structure
162
163	```json
164	{
165	"frontmatter": {
166	"name": "string (required)",
167	"description": "string (required)",
168	"allowed-tools": "string (optional)",
169	"model": "enum (optional)",
170	"agentType": "agent (optional)"
171	},
172	"content": "string (markdown with H1, persona, instructions)"
173	}
174	```
175
176	### Common Validation Errors
177
178	\| Error \| Cause

People who liked this also viewed

6 picks

Type