Generate images via relay endpoints with saved prompts and metadata sidecars.
npx skills add https://github.com/AwakeFantasy/relay-imagegen --skill relay-imagegen
Please help me install this Agent Skill. Skill: relay-imagegen Source: https://github.com/AwakeFantasy/relay-imagegen/tree/main Command: npx skills add https://github.com/AwakeFantasy/relay-imagegen --skill relay-imagegen
Generate or edit images through a local OpenAI-compatible relay or proxy endpoint, with saved prompt files and non-secret run metadata. Use when you want to avoid persistent OPENAI_API_KEY environment variables and reuse current Codex/ccswitch configurations.
Generate an image through a relay-based OpenAI-compatible endpoint using prompts from prompts/*.txt and save a non-secret run metadata sidecar in generated/.
High-intent search queries and answers about installing and using this skill.
Image generation through a local relay endpoint saves prompts under prompts/*.txt and writes a non-secret .meta.json sidecar with run metadata next to each generated image. It reads Codex or ccswitch settings, injects API keys only into the child process, and validates output dimensions automatically.
Yes, relay-based image generation avoids persistent OPENAI_API_KEY environment variables by injecting API keys only into the child process. It reads existing Codex, ccswitch configurations, or private JSON configs to authenticate, ensuring keys are not left exposed in your environment.
This image generation workflow supports multiple configuration sources including Codex, ccswitch, and private JSON configs. It reads these settings to determine relay endpoints and API keys, enabling flexible workflows without needing separate manual setup for each run.
The best way to reuse prompts for relay image generation is saving them under prompts/*.txt files. This centralizes prompts for consistency and reuse, allowing the workflow to read them directly and write a non-secret run metadata sidecar capturing prompt snapshots for each generated image.
To save non-secret run metadata sidecars, the workflow automatically writes a .meta.json file next to each generated image in the generated/ directory. This sidecar captures run metadata and prompt snapshots without exposing secret API keys, ensuring safe version tracking.
Limitations of using private JSON configs for relay image generation include requiring valid relay endpoints and correct output dimensions. The workflow validates output dimensions and relies on private JSON configs or Codex settings, meaning misconfigured endpoints will cause generation failures.
ktkarchiveGenerate and edit images from text prompts and reference images via OpenAI Codex OAuth.
SJTU-IPADSAutomate deterministic image editing tasks on image collections with Pillow.
albibenniConfigure ESLint, Prettier, and TypeScript settings based on detected frameworks and package managers.
manashmandalGenerate images from text prompts using OpenAI image APIs and extract OCR text from images.
