원클릭으로
create-skill
Create opencode skills with correct structure, frontmatter, naming, and placement
Codex 또는 Claude로 설치 이 Prompt를 복사해 Codex, Claude 또는 다른 어시스턴트에 붙여 넣으면 Skill 페이지를 검토하고 설치를 진행할 수 있습니다.
메뉴
Create opencode skills with correct structure, frontmatter, naming, and placement
Codex 또는 Claude로 설치 이 Prompt를 복사해 Codex, Claude 또는 다른 어시스턴트에 붙여 넣으면 Skill 페이지를 검토하고 설치를 진행할 수 있습니다.
SOC 직업 분류 기준
| name | create-skill |
| description | Create opencode skills with correct structure, frontmatter, naming, and placement |
Create opencode skills following these rules.
https://opencode.ai/docs/skills.md
config/opencode/skills/<name>/SKILL.md (nix-managed, symlinked to ~/.config/opencode/skills/<name>/SKILL.md).
OpenCode also scans .opencode/skills/, .claude/skills/, .agents/skills/ walking up from cwd to git root. Global: ~/.config/opencode/skills/, ~/.claude/skills/, ~/.agents/skills/.
YAML frontmatter required:
name — must match directory namedescription — max 1024 chars, specific enough for agent selectionOptional: license, compatibility, metadata (string map). Unknown fields ignored.
^[a-z0-9]+(-[a-z0-9]+)*$Skills appear in skill tool description as <available_skills> list (name + description). Agent calls skill({ name: "skill-name" }) to load full content into context.
Trust precise names over explanations. Flat lists over nesting.
Pattern-based, configured in opencode.json:
allow — loads immediatelydeny — hidden from agentask — user approves before loading{
"permission": {
"skill": {
"*": "allow",
"internal-*": "deny",
"experimental-*": "ask"
}
}
}
Custom agents (frontmatter): permission: { skill: { "pattern": "allow" } }
Built-in agents (opencode.json): agent: { "agent-name": { permission: { skill: { "pattern": "allow" } } } }
Custom agents: tools: { skill: false }
Built-in agents: agent: { "agent-name": { tools: { skill: false } } }
Omits <available_skills> section entirely.
SKILL.md all capsname + description