codex-imagelisted

# Codex Image Generation Generate an image by delegating to the `ask-codex` MCP tool with a prompt-engineered template that triggers Codex's built-in `image_generation` tool. The result is saved to disk and the path is returned to the user. ## Prerequisites - `codex-cli` >= 0.125.0 installed and authenticated - `image_generation` feature flag enabled (default: stable + true). Verify with `codex features list | grep image_generation` - The `ask-codex` MCP tool available (from `ask-codex-mcp` or the `ask-llm-mcp` orchestrator) ## Instructions ### Phase 1: Build the image prompt Take the user's natural-language ask and prepare it for image generation. Rules: - **Keep the user's intent intact** — do not change meaning. Refine, don't replace. - **If the prompt is sparse** (e.g., "an apple") AND the surrounding conversation has design context (e.g., a LinkedIn post, an article), enrich with that context: aspect ratio, style cues, what to avoid. - **If the prompt is already detailed**, send it through verbatim — the user knows what they want. - **Default exclusions worth adding**: "no humanoid figures, no glowing brains, no chatbot iconography" if the topic is AI / LLM-related (these are diffusion-model failure modes and the user will almost certainly want them excluded). If you significantly enrich the prompt, briefly tell the user what you added and let them push back before dispatching. ### Phase 2: Determine output path Convention: - **Default path**: `/tmp/codex-imag