菜单
Use when the user asks to generate, create, make, or save images with xAI, Grok Imagine, or a Grok-connected xAI image generation tool.
用 Codex 或 Claude 帮你安装 复制这段 Prompt，粘贴到 Codex、Claude 或其他助手里，让它检查 Skill 页面并帮你完成安装。
基于 SOC 职业分类
| name | xai_image_generate |
| description | Use when the user asks to generate, create, make, or save images with xAI, Grok Imagine, or a Grok-connected xAI image generation tool. |
Use this skill for xAI/Grok Imagine image requests through the Grok-connected xAI image-generation tool.
search_tool to discover the available tool schema first:{
"query": "xai image generate"
}
{
"prompt": "tiny minimalist blue dot icon on white background",
"resolution": "2k",
"artifact_dir": ".grok-xai-artifacts"
}
For base64 image payloads saved locally, request the response format supported by the tool schema and pass an artifact directory when available.
prompt: required image prompt.model: defaults are tool-owned unless the user asks for a specific model.n: number of images.resolution: commonly 1k or 2k when supported.response_format: url or b64_json when supported.size: provider-specific size string.artifact_dir: local directory for saved image artifacts, when supported.When artifact_dir is present, return the upstream JSON plus artifact metadata. Saved artifacts should include a local path, mime_type, source, and status when the tool provides them. If a remote URL cannot be downloaded, preserve its upstream URL instead of hiding it.
Use existing Grok/xAI credentials exposed by the host or MCP server. Never print, quote, summarize, or store OAuth tokens, refresh tokens, bearer tokens, or API keys.
Execute a Prometheus work plan in GrokBuild with `/goal` state, Boulder state, evidence ledger updates, worktree discipline, parallel subagents, and explicit continuation. Use after planning when the user says start work, execute plan, continue plan, resume plan, or asks to run a .omo/plans plan.
Goal-like loop that uses ultrawork mode to decompose work into systematic, evidence-bound steps.
MUST USE for planning before coding: 5+ steps, ambiguous scope, multiple modules, architecture decisions, a vague 'just make it good / figure out what to build' brief, or any request to plan, interview, or break work down. Explore-first planning consultant (Prometheus) that grounds in the codebase, asks only the forks exploration cannot resolve - or researches them to best practice when the intent is fuzzy - waits for explicit approval, then writes ONE decision-complete work plan a worker executes with zero further interview. Triggers: ulw-plan, plan this, make a plan, plan before coding, interview me, break this down, start planning, plan mode, just make it good, figure out what to build.
Execute a Prometheus work plan in GrokBuild with `/goal` state, Boulder state, evidence ledger updates, worktree discipline, parallel subagents, and explicit continuation. Use after planning when the user says start work, execute plan, continue plan, resume plan, or asks to run a .omo/plans plan.
Goal-like loop that uses ultrawork mode to decompose work into systematic, evidence-bound steps.
Persistent code-mode kernels (JS/Python) for incremental computation — omp/senpi eval analog on GrokBuild via lfg MCP. Use when multi-step code execution, data exploration, or stateful REPL beats one-shot bash.