메뉴
Primary fan-out orchestrator for coordinated work across the Cerid family of repositories. Optimized for parallel subagent usage rather than sequential execution.
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
Project-specific UI/UX constraints for **Cerid AI repos only** (cerid-ai-internal, cerid-ai, cerid-trading-agent, cerid-boardroom, cerid-* client repos, and any future cerid-derivative repo). Use ONLY when the current working directory is inside such a repo. Composes with global design-intelligence skills (e.g. ui-ux-pro-max) so generic recommendations land on-brand. Covers the stack pin (React 19 + shadcn/ui new-york + Tailwind v4 + lucide), brand voice + anti-patterns (no AI purple/pink gradients, no marketing funnel), the canonical tier-colour system (private mode + verification bands), the 4-state UX matrix required per pane, the active design-drift gate + allowlist, the canonical primitive ladder under src/web/src/components/ui/, the cross-pane navigation contract, and the accessibility floor.
Primary orchestrator for operating across the Cerid AI family of repositories. Use when the user wants coordinated work across multiple cerid-* repos (status, analysis, refactoring, audits, etc.). This skill is optimized for Grok's subagent model — it decides when to fork parallel Explore/Researcher/Reviewer subagents instead of doing sequential work.
High-level system orchestrator for the entire Cerid AI ecosystem. Use for complex, multi-domain tasks that require routing between different areas (architecture, KB, trading, frontend, infrastructure, cross-repo work). This skill understands the full system map and decides which specialized skills, commands, or subagents to invoke.
Specialized subagent for enforcing Cerid AI architecture rules (core/app separation, import-linter, layering, preservation invariants). Must be used before and during any structural refactoring or architecture changes in cerid-ai, cerid-ai-internal, or related repos. Can be invoked as a dedicated reviewer persona.
Project-specific UI/UX and design system constraints for Cerid AI repos only. Grok-optimized version. Use when working inside cerid-ai, cerid-ai-internal, cerid-trading-agent, cerid-boardroom, cerid-* client repos, or any future cerid-derivative. Composes with global design skills. Covers the stack pin (React 19 + Tailwind + shadcn/ui + specific component patterns), voice of the product, and brand constraints.
Use when the user asks what's loaded in the current Grok Build session, why context feels heavy, how many tokens subagents are using, or how to reduce overhead. Covers Grok-specific concepts: injected system prompt size, available skills (from ~/.grok/skills + project .grok/skills + bundled), active MCP servers (grok_com_github etc.), subagent types (implement, review, explore, plan, best-of-n), and the workspace path. Recommends trims. Complements (does not duplicate) the Claude `audit-context` skill.
| name | grok-across-cerid (card) |
| description | Primary fan-out orchestrator for coordinated work across the Cerid family of repositories. Optimized for parallel subagent usage rather than sequential execution. |
Quick Guidance:
explore/researcher subagents in parallel.Fork the full version from ~/Develop/.grok/skills/family/grok-across-cerid/ for complex cross-repo work.