create-agent
About
This skill creates a new agent definition file using the agent-almanac template and conventions. It guides developers through persona design, tool selection, skill assignment, and registry integration. Use it when adding a specialized agent to the library or defining a domain-specific assistant with curated capabilities.
Quick Install
Claude Code
Recommendednpx skills add pjt222/agent-almanac -a claude-code/plugin add https://github.com/pjt222/agent-almanacgit clone https://github.com/pjt222/agent-almanac.git ~/.claude/skills/create-agentCopy and paste this command in Claude Code to install this skill
Documentation
造代
定 Claude Code 子代之身。含精目、選具、賦技、備檔。
用
- 為未涵域加專代
- 化復流或模為可重之代
- 以選技具造域專助
- 分過泛代為專責代
- 組多代團前設新員
入
- 必:代名(小寫、kebab-case,如
data-engineer) - 必:一行述代主目
- 必:解所治問之目述
- 可:模選(默:
sonnet;代:opus、haiku) - 可:優先(默:
normal;代:high、low) - 可:
skills/_registry.yml中之技列 - 可:代需 MCP 服(如
r-mcptools、hf-mcp-server)
行
一:設代身
擇明專之代身:
- Name:小寫 kebab-case、述角。起於名詞或域辭:
security-analyst、r-developer、tour-planner。避泛名如helper、assistant - Purpose:一段述此代解之具問。問:「此代為何無存代可代?」
- Communication style:考域。技代宜精引重。創代可更探。合規代宜正式而審導
繼前察 53 代之重:
grep -i "description:" agents/_registry.yml | grep -i "<your-domain-keywords>"
得: 無存代涵同隙。若存代部重→擴之非新建。
敗: 重代存→擴彼技列或窄新代範為補非復。
二:選具
擇代需之最小具集。最少特權原則:
| 具集 | 用時 | 例代 |
|---|---|---|
[Read, Grep, Glob] | 讀析、評、審 | code-reviewer、security-analyst、auditor |
[Read, Grep, Glob, WebFetch] | 析加外查 | senior-researcher |
[Read, Write, Edit, Bash, Grep, Glob] | 全發——建/改碼 | r-developer、web-developer、devops-engineer |
[Read, Write, Edit, Bash, Grep, Glob, WebFetch, WebSearch] | 發加外研 | polymath、shapeshifter |
僅析碼者勿含 Bash。勿含 WebFetch 或 WebSearch 除代實需外查。
得: 具列僅含代實用者。
敗: 評能——若能不需具→除具。
三:選模
按任複擇模:
sonnet(默):多代。推理與速衡。用於發、評、析、標流opus:複推、多步謀、細判。用於高級代、架決、深域專haiku:簡速應。用於直查、式、模填
得: 模合代主用例之認知需。
敗: 疑則用 sonnet。測揭推不足方升 opus。
四:賦技
覽技庫選代域相關技:
# List all skills in a domain
grep -A3 "domain-name:" skills/_registry.yml
# Search for skills by keyword
grep -i "keyword" skills/_registry.yml
為 frontmatter 築技列:
skills:
- skill-id-one
- skill-id-two
- skill-id-three
要:諸代自動繼庫級 default_skills 之默技(meditate、heal)。勿於代 frontmatter 列此二除非為代法核(如 mystic 代列 meditate 因冥助為其主)。
得: 技列含 3-15 存於 skills/_registry.yml 之技 ID。
敗: 驗技 ID 存:grep "id: skill-name" skills/_registry.yml。除不合者。
五:書代檔
複模填 frontmatter:
cp agents/_template.md agents/<agent-name>.md
填 YAML frontmatter:
---
name: agent-name
description: One to two sentences describing primary capability and domain
tools: [Read, Write, Edit, Bash, Grep, Glob]
model: sonnet
version: "1.0.0"
author: Philipp Thoss
created: YYYY-MM-DD
updated: YYYY-MM-DD
tags: [domain, specialty, relevant-keywords]
priority: normal
max_context_tokens: 200000
skills:
- assigned-skill-one
- assigned-skill-two
# Note: All agents inherit default skills (meditate, heal) from the registry.
# Only list them here if they are core to this agent's methodology.
# mcp_servers: [] # Uncomment and populate if MCP servers are needed
---
得: YAML frontmatter 無誤解析。諸必欄(name、description、tools、model、version、author)存。
敗: 驗 YAML 文法。常誤:版串缺引、縮進誤、具列括未閉。
六:書目與能
代模之位:
Purpose:一段述此代解之具問與值。具——名域、流、果。
Capabilities:粗體引項。代能多→按類組:
## Capabilities
- **Primary Capability**: What the agent does best
- **Secondary Capability**: Additional functionality
- **Tool Integration**: How it leverages its tools
Available Skills:各賦技含短述。用裸技 ID(斜命名):
## Available Skills
- `skill-id` - Brief description of what the skill does
得: 目具(非「助發」)、能具驗、技列合 frontmatter。
敗: 目感泛→答:「用者當請此代作何具任?」以此為目。
七:書用例與例
予 2-3 用例顯如何召代:
### Scenario 1: Primary Use Case
Brief description of the main scenario.
> "Use the agent-name agent to [specific task]."
### Scenario 2: Alternative Use Case
Description of another common use case.
> "Spawn the agent-name to [different task]."
加 1-2 具例顯用請與期代行:
### Example 1: Basic Usage
**User**: [Specific request]
**Agent**: [Expected response pattern and actions taken]
得: 例實、顯實值、召式合 Claude Code 規。
敗: 心試例——代實可以賦具技成請乎?
八:書限與參
Limitations:3-5 誠限。代不能、不當、或果差之處:
## Limitations
- Cannot execute code in language X (no runtime available)
- Not suitable for tasks requiring Y — use Z agent instead
- Requires MCP server ABC to be running for full functionality
See Also:引補代、相關導、相關團:
## See Also
- [complementary-agent](complementary-agent.md) - handles the X side of this workflow
- [relevant-guide](../guides/guide-name.md) - background knowledge for this domain
- [relevant-team](../teams/team-name.md) - team that includes this agent
得: 限誠具。See Also 引存檔。
敗: 察引檔存:ls agents/complementary-agent.md。
九:加於庫
編 agents/_registry.yml 於字母位加新代:
- id: agent-name
path: agents/agent-name.md
description: Same one-line description from frontmatter
tags: [domain, specialty]
priority: normal
tools: [Read, Write, Edit, Bash, Grep, Glob]
skills:
- skill-id-one
- skill-id-two
增檔首 total_agents 計。
得: 庫項合代檔 frontmatter。total_agents 等實代項數。
敗: 以 grep -c "^ - id:" agents/_registry.yml 計項、驗合 total_agents。
十:驗發現
Claude Code 自 .claude/agents/ 發現代。此庫中此目為 agents/ 之軟連:
# Verify the symlink exists and resolves
ls -la .claude/agents/
readlink -f .claude/agents/<agent-name>.md
軟連全則無須外動——新代檔自動可發現。
行 README 自更:
npm run update-readmes
得: .claude/agents/<agent-name>.md 解至新代檔。agents/README.md 含新代。
敗: 軟連破→重建:ln -sf ../agents .claude/agents。npm run update-readmes 敗→察 scripts/generate-readmes.js 存且 js-yaml 裝。
十一:架譯
諸代必。此步施於人作者與循此程之 AI 代。勿略——缺譯積為陳備。
承新代後即為諸 4 支 locales 架譯檔:
for locale in de zh-CN ja es; do
npm run translate:scaffold -- agents <agent-name> "$locale"
done
續譯各檔之架詞(碼塊與 ID 留英)。終重生態檔:
npm run translation:status
得: i18n/{de,zh-CN,ja,es}/agents/<agent-name>.md 建四檔,source_commit 皆合現 HEAD。npm run validate:translations 顯零陳警於新代。
敗: 架敗→驗代存於 agents/_registry.yml。態檔不更→顯行 npm run translation:status——CI 不自觸。
驗
- 代檔存於
agents/<agent-name>.md - YAML frontmatter 無誤解析
- 諸必欄存:
name、description、tools、model、version、author -
name合檔名(無.md) - 諸節存:Purpose、Capabilities、Available Skills、Usage Scenarios、Examples、Limitations、See Also
- Frontmatter 中技存於
skills/_registry.yml - 默技(
meditate、heal)非列除非為代法核 - 具列循最少特權
- 代於
agents/_registry.yml含正路與合備 - 庫中
total_agents計已更 -
.claude/agents/軟連解至新代檔 - 無顯著重於存代
忌
- 具過授:僅讀析而含
Bash、Write、WebFetch→破最少特權致副效。始於最小集、能需方加 - 缺或誤技賦:列庫無之技 ID 或全忘賦技。加前以
grep "id: skill-name" skills/_registry.yml驗 - 無謂列默技:加
meditate或heal於代 frontmatter 而庫已繼。僅核方列(如mystic、alchemist、gardener、shaman) - 範重存代:建代復 53 存代之能。先搜庫、考擴存代之技
- 目能泛:書「助發」而非「架 R 包含全構、備、CI/CD 設」。具為代用與可發之源
參
create-skill- 建 SKILL.md 而非代檔之並程create-team- 組多代為調團(計中)commit-changes- 承新代檔與庫更
GitHub Repository
Related Skills
content-collections
MetaThis skill provides a production-tested setup for Content Collections, a TypeScript-first tool that transforms Markdown/MDX files into type-safe data collections with Zod validation. Use it when building blogs, documentation sites, or content-heavy Vite + React applications to ensure type safety and automatic content validation. It covers everything from Vite plugin configuration and MDX compilation to deployment optimization and schema validation.
polymarket
MetaThis skill enables developers to build applications with the Polymarket prediction markets platform, including API integration for trading and market data. It also provides real-time data streaming via WebSocket to monitor live trades and market activity. Use it for implementing trading strategies or creating tools that process live market updates.
creating-opencode-plugins
MetaThis skill helps developers create OpenCode plugins that hook into 25+ event types like commands, files, and LSP operations. It provides the plugin structure, event API specifications, and implementation patterns for JavaScript/TypeScript modules. Use it when you need to intercept, monitor, or extend the OpenCode AI assistant's lifecycle with custom event-driven logic.
sglang
MetaSGLang is a high-performance LLM serving framework that specializes in fast, structured generation for JSON, regex, and agentic workflows using its RadixAttention prefix caching. It delivers significantly faster inference, especially for tasks with repeated prefixes, making it ideal for complex, structured outputs and multi-turn conversations. Choose SGLang over alternatives like vLLM when you need constrained decoding or are building applications with extensive prefix sharing.
