원클릭으로
agent-creator
Use when creating expert agents. Generates an agent .toml with the required fields, attached skills, and required sections.
Codex 또는 Claude로 설치 이 Prompt를 복사해 Codex, Claude 또는 다른 어시스턴트에 붙여 넣으면 Skill 페이지를 검토하고 설치를 진행할 수 있습니다.
메뉴
Use when creating expert agents. Generates an agent .toml with the required fields, attached skills, and required sections.
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
Systematic development workflow: Analyze → Plan → Execute → eLicit → eXamine. Use for ANY development task: features, bug fixes, refactoring, hotfixes. Triggers: "implement", "create", "build", "fix", "add feature", "refactor", "develop". Auto-detects project type (Laravel, Next.js, React, Swift) and loads framework-specific references. Enforces: files <100 lines, interfaces separated, SOLID principles, expert self-review, sniper validation. Modes: --auto (default), --manual, --skip-elicit
Auto-review skill for expert agents. After coding, expert applies elicitation techniques to self-correct before sniper validation. Inspired by BMAD-METHOD. Use when: an expert agent needs to self-review and self-correct code after the Execute phase, before sniper validation.
Use when marking a task as complete, finishing a feature, or claiming a bug is fixed. Ensures functional resolution is verified with evidence before closing. Do NOT use for: lint/type/code-quality validation (use code-quality / sniper AFTER functional verification passes).
Smart conventional commit with security validation, branch flow enforcement, and auto-detection. Use when committing changes, saving work, staging and committing, or choosing a conventional commit message.
Use when committing, branching, opening PRs, or deciding merge strategy. Covers GitHub Flow (default), trunk-based, branch naming conventions, squash vs rebase, branch lifecycle, and protected branch enforcement.
Universal post-commit actions. CHANGELOG update for all repos (git tag is created POST-MERGE, not here — see `commit` command Step 8). Plugin version bumping for marketplace repos. Triggered after any code commit (except wip/amend/undo).
| name | agent-creator |
| description | Use when creating expert agents. Generates an agent .toml with the required fields, attached skills, and required sections. |
Before ANY agent creation, spawn 2 subagents in parallel (one dispatch, spawn_agent). Generated agents reference other agents by their bare name via spawn_agent too, so this same pattern is what they must use internally:
mcp__context7__query-docs is a direct MCP call, not a spawned agent — invoke it directly (alongside the 2 subagents) to get examples from existing agents.
After creation, run sniper for validation.
| Action | When to Use |
|---|---|
| New Agent | New domain/framework expert needed |
| Adapt | Copy from similar agent (Next.js → React) |
| Update | Add skills, modify sections |
name, description, model, model_reasoning_effort, sandbox_mode, developer_instructions, [[skills.config]]developer_instructions)solid-<stack> skillhooks/hooks.json - Plugin level, never in the agent .toml## Output Format section (status, files_changed, errors, sources) — an agent invoked by a lead must return structured data, not proseplugins/<plugin-name>/
├── agents/
│ └── <agent-name>.toml # Agent definition (TOML)
├── skills/
│ ├── skill-a/
│ └── solid-[stack]/
├── hooks/
│ └── hooks.json # Pre/Post-tool validation (plugin level)
├── scripts/
│ └── validate-*.sh # Hook scripts
└── .codex-plugin/
└── plugin.json
→ See architecture.md for details
| Topic | Reference | When to Consult |
|---|---|---|
| Architecture | architecture.md | Understanding agent structure |
| Frontmatter | frontmatter.md | TOML field configuration |
| Required Sections | required-sections.md | Mandatory content |
| Hooks | hooks.md | Pre/Post tool validation |
| Registration | registration.md | Plugin manifest |
| Template | When to Use |
|---|---|
| agent-template.md | Creating new agent |
| hook-scripts.md | Validation scripts |
# 1. Research existing agents
→ explore-codebase + research-expert
# 2. Create files
touch plugins/<plugin>/agents/<agent-name>.toml
touch plugins/<plugin>/scripts/validate-<stack>-solid.sh
chmod +x plugins/<plugin>/scripts/*.sh
# 3. Register in the plugin manifest
# 4. Validate
→ sniper
# 1. Copy similar agent
cp plugins/nextjs-expert/agents/nextjs-expert.toml plugins/new-plugin/agents/new-expert.toml
# 2. Adapt with sed
sed -i '' "s/nextjs/newstack/g; s/Next\.js/NewStack/g" agents/new-expert.toml
# 3. Update skills, model, register
name, description, model, model_reasoning_effort, sandbox_mode, developer_instructions)solid-[stack], attached via [[skills.config]]hooks/hooks.jsonWhen creating an agent, you often need to create skills too.
Use the skill-creator skill ($skill-creator) to create skills for the agent:
| Scenario | Action |
|---|---|
| New agent needs skills | Create skills with skill-creator first |
| Agent references skills | Ensure skills exist in skills/ |
| Adapting agent | Adapt related skills too |
solid-[stack] skill for SOLID ruleshooks/hooks.json$PLUGIN_ROOT).toml