Agent Development for Claude Code Plugins

Agent Development for Claude Code Plugins

Overview

Agents are autonomous subprocesses that handle complex, multi-step tasks independently. Each agent is a markdown file in the agents/ directory with YAML frontmatter defining its configuration and a markdown body serving as its system prompt.

Agent File Format

---
name: agent-identifier
description: |
  Use this agent when [triggering conditions].

  <example>
  Context: [Situation]
  user: "[User request]"
  assistant: "[How to respond and invoke agent]"
  <commentary>
  [Why this agent should trigger]
  </commentary>
  </example>

model: inherit
color: blue
tools:
  - Read
  - Write
  - Grep
---

System prompt body in second person ("You are...")

Frontmatter Fields Reference

name (required)

Agent identifier for namespacing and invocation.

Rule	Detail
Length	3-50 characters
Format	Lowercase letters, numbers, hyphens only
Start/end	Must be alphanumeric (not hyphen)
Convention	Role-based: code-reviewer, test-generator, domain-expert

Invalid names: ag (too short), -agent- (starts/ends with hyphen), my_agent (underscores)

description (required - most critical field)

Defines WHEN Claude should trigger this agent. Poor descriptions = agent never triggers.

Must include:

1. Triggering conditions ("Use this agent when...")
2. 2-4 <example> blocks showing usage scenarios
3. Each example: context, user request, assistant response, commentary
4. Both proactive and reactive triggering scenarios

Good description pattern:

description: |
  Use this agent when the user needs help with [domain]. Trigger for:
  - [Scenario 1]
  - [Scenario 2]
  - [Scenario 3]

  <example>
  Context: [Specific situation]
  user: "[What user says]"
  assistant: "[How Claude responds and invokes agent]"
  <commentary>
  [Why this is the right agent for this request]
  </commentary>
  </example>

Common mistake: Vague descriptions without examples. "Helps with code review" will rarely trigger. Include concrete examples with exact user phrases.

Example block rules:

• Keep example blocks concise — assistant response should be 1-2 sentences, not full code
• Limit to 3-7 example blocks total (more dilutes matching quality)
• Do NOT include full JSON schemas, code samples, or CLI output in examples
• Examples show when to trigger and how to respond, not the domain content itself

Skill coverage requirement (CRITICAL): When the agent delegates to skills, every skill MUST have at least one <example> block that would route to it. Count skills, count examples, and verify full coverage. If the plugin has 9 skills and only 7 trigger examples, 2 skills will have reduced activation reliability. Add examples until every skill has explicit coverage. When there are more skills than the 7-example limit allows, combine related skills into shared examples that mention both domains.

model (required)

Value	When to use
inherit	Default choice - uses parent session's model
sonnet	Balanced capability/speed
opus	Most capable, for complex reasoning
haiku	Fast/cheap, for simple validation

Always use inherit unless the agent specifically needs a different capability level.

color (required)

Visual identifier in UI. Choose based on agent function:

Color	Use for
blue / cyan	Analysis, review, research
green	Success-oriented, generation, creation
yellow	Caution, validation, checking
red	Critical, security, destructive operations
magenta	Creative, design, architecture

Use distinct colors for different agents within the same plugin.

tools (optional)

Restrict agent to specific tools. Principle of least privilege - only grant what's needed.

# Read-only analysis
tools: ["Read", "Grep", "Glob"]

# Code generation
tools: ["Read", "Write", "Edit", "Grep", "Glob"]

# Full access (omit field entirely)
# tools: (not specified)

Common tool names: Read, Write, Edit, Grep, Glob, Bash, WebSearch, WebFetch, Skill, Agent

MCP tools use format: mcp__server-name__tool-name

System Prompt Design

The markdown body becomes the agent's system prompt. Write in second person ("You are...", "You will...").

Structure Template

You are [role] specializing in [domain].

## Core Responsibilities
1. [Primary responsibility]
2. [Secondary responsibility]

## Process
1. [Step one]
2. [Step two]
3. [Step three]

## Quality Standards
- [Standard 1]
- [Standard 2]

## Output Format
- [What to include]
- [How to structure results]

## Edge Cases
- [Situation]: [How to handle]

Best Practices

DO:

• Write in second person ("You are...", "You will...")
• Be specific about responsibilities and process steps
• Define output format clearly
• Address edge cases
• Include skill activation instructions if the agent should load skills
• Keep the agent body as a lean orchestrator (see below)

DON'T:

• Write in first person ("I am...", "I will...")
• Be vague or generic ("help with stuff")
• Skip process steps
• Leave output format undefined
• Omit quality standards
• Embed domain knowledge that belongs in skills (see below)

Lean Orchestrator Pattern (CRITICAL)

An agent body must be a lean orchestrator, NOT a domain knowledge dump. The agent delegates to skills for detailed knowledge.

Agent Body Size Limits

Metric	Target	Hard Maximum
Word count	1,500-2,500 words	3,000 words
Character count	~10,000-15,000 chars	20,000 chars

What Belongs in the Agent Body

Section	Required	Purpose
Role identity	Yes	"You are [role] specializing in [domain]"
Skill activation rules	Yes	Topic-to-skill mapping table
High-level process	Yes	Design/workflow steps
Output format	Yes	What to include in responses
Brief service summaries	Optional	2-3 sentences per area to help decide which skill to load
Edge cases / troubleshooting tips	Optional	Quick reference only

What Does NOT Belong in the Agent Body

• Detailed domain knowledge — belongs in skills
• Complete CLI/API references — belongs in skill references/
• Full code examples — belongs in skill examples/
• Duplicated skill content — if it's in a skill, do NOT repeat it in the agent

Anti-Pattern: Content Duplication

NEVER duplicate content between the agent body and skills. This is the most common mistake and causes massive context bloat.

Bad: Agent body contains a full "Plugin.json Schema" section AND the plugin-master skill also contains it. Good: Agent body says "For plugin.json schema details, load plugin-master:plugin-master" and keeps only a 1-sentence summary.

Lean Orchestrator Template

You are [role] specializing in [domain].

## Skill Activation - CRITICAL
[Topic-to-skill mapping table — this is the heart of the agent]

## Core Responsibilities
[2-5 bullet points on what this agent does]

## Process
[5-7 step workflow for handling user requests]

## Quality Standards
[Brief checklist — 5-10 items]

## Output Format
[What to include in responses]

Description Size Limits

Agent descriptions should be concise and effective:

Element	Guideline
Intro text	1-2 sentences on when to trigger
Example blocks	3-7 blocks covering diverse scenarios
Total description	Should fit naturally — focus on quality trigger examples over length

Agent Design Principles (2025)

Agent-First Plugin Design

• Primary plugin interface is ONE expert agent named {domain}-expert
• Plugin named docker-master → agent named docker-expert
• Only 0-2 slash commands for automation workflows
• Users interact conversationally, not through command menus

Single Responsibility

Each agent should have a clear, focused purpose. Don't create "do everything" agents. If a plugin needs multiple capabilities, use one expert agent that loads different skills based on context.

Skill Integration

Expert agents should load relevant skills before answering. Include skill activation instructions in the system prompt:

## Skill Activation
When the user asks about [topic], load `plugin-name:skill-name` before responding.

Preventing Trigger Phrase Overlap Between Skills

When a plugin has multiple skills, their trigger phrases and description terms must not create ambiguity. If two skills both claim the same keyword (e.g., both "programmatic-development" and "tmdl-mastery" claim "TMDL"), the agent cannot reliably route requests.

Disambiguation rules:

1. Audit trigger terms across all skills — list every trigger phrase from every skill description side by side. Flag any term that appears in more than one skill.
2. Assign exclusive ownership — each ambiguous term must belong to exactly one skill. The other skill should use a more specific phrase (e.g., "TMDL file editing" vs. "programmatic deployment using TMDL").
3. Add disambiguation hints to the agent's skill activation table — for terms that could route to multiple skills, add a clarifying note: "TMDL editing/syntax → tmdl-mastery; TMDL in deployment pipelines → programmatic-development".
4. Test with ambiguous queries — after writing descriptions, mentally test phrases like "help me with TMDL" and verify the routing is unambiguous.

Validation Checklist

Before finalizing an agent:

• Name: 3-50 chars, lowercase, hyphens, starts/ends alphanumeric
• Description: includes triggering conditions and 2-4 <example> blocks
• Every skill has trigger coverage: count skills and verify each has at least one example that routes to it
• No trigger phrase overlap: no ambiguous keyword claimed by multiple skills without disambiguation
• Model: set to inherit (unless specific need)
• Color: appropriate for agent function
• Tools: restricted to minimum needed (or omitted for full access)
• System prompt: second person, clear responsibilities, defined process and output
• Frontmatter: valid YAML with all required fields
• File location: agents/agent-name.md

Testing

1. Write agent with specific triggering examples
2. Use similar phrasing to examples in your test queries
3. Verify Claude loads the agent for matching requests
4. Test that the agent follows its defined process
5. Check output matches defined format
6. Test edge cases mentioned in system prompt

Common Mistakes

Mistake	Fix
Vague description without examples	Add 2-4 <example> blocks with concrete user phrases
Skills without trigger examples	Every skill must have at least one example that routes to it — count and verify
Trigger phrase overlap between skills	Audit all skill descriptions for shared keywords; assign exclusive ownership or add disambiguation
model: sonnet when inherit works	Use inherit unless agent needs specific capability
Too many tools granted	Restrict to minimum needed tools
Generic system prompt	Be specific about process, output format, quality standards
No skill activation	Add skill loading instructions for knowledge-dependent agents
Multiple agents in one plugin	Use one expert agent with skills for different topics
Example blocks with full code/JSON	Keep examples concise (1-2 sentence responses); code belongs in skills
Same cross-cutting block in every skill	Put platform guidelines in agent body or one shared reference, not each SKILL.md