← Back to Skills Marketplace
Baoyu Image Gen
by
Jim Liu 宝玉
· GitHub ↗
· v2.1.0
· MIT-0
4168
Downloads
1
Stars
92
Active Installs
4
Versions
Install in OpenClaw
/install baoyu-image-gen
Description
AI image generation with OpenAI GPT Image 2, Azure OpenAI, Google, OpenRouter, DashScope, Z.AI GLM-Image, MiniMax, Jimeng, Seedream, Replicate and Agnes APIs...
Usage Guidance
Install only if you are comfortable with external image providers receiving your prompts and reference images. Avoid the `codex-cli` provider unless you intentionally want to use your logged-in Codex session and accept that it runs a full-access local Codex subprocess; prefer direct API providers when possible, avoid sensitive local reference images, and do not set wrapper/base-url override environment variables unless you fully trust the target.
Capability Assessment
Purpose & Capability
Most provider API calls, credential use, reference-image uploads, and local output writes are coherent with an image-generation skill. The material exception is the codex-cli provider, which invokes `codex exec --sandbox danger-full-access`, creating a high-impact local agent path for a media-generation task.
Instruction Scope
The skill discloses that codex-cli is explicitly selected and not auto-selected, and its wrapper instructs Codex to use only image generation. However, arbitrary user prompt text is still passed into a full-access agent subprocess, so the practical authority is broader than the stated image-only workflow.
Install Mechanism
No hidden install-time execution was found. The skill requires `bun` or `npx`, API keys or Codex login depending on provider, and documents a `BAOYU_CODEX_IMAGEGEN_BIN` override that can execute a user-specified wrapper path.
Credentials
Network calls to image providers and sending prompts/reference images are expected for this purpose, but full local Codex execution and unvalidated provider-returned image downloads are higher-risk than direct scoped API calls.
Persistence & Privilege
The skill writes outputs and configuration as expected, but the Codex path uses the user's logged-in Codex session, writes prompt-related temporary files, and always persists raw Codex stdout/stderr logs under a temp directory; optional cache and diagnostic logs add more retained local data.
How to Use
- Make sure OpenClaw is installed (local or Docker)
- Run the install command in chat:
/install baoyu-image-gen - After installation, invoke the skill by name or use
/baoyu-image-gen - Provide required inputs per the skill's parameter spec and get structured output
Version History
v2.1.0
baoyu-image-gen 2.1.0
- Added support for new image providers: Codex CLI and Agnes (see new references and scripts).
- Expanded provider options to include `codex-cli` and `agnes`; `codex-cli` requires explicit selection.
- Introduced batch file building script (`scripts/build-batch.ts`) for advanced batch generation workflows.
- Additional references and documentation covering new providers and usage examples.
- Internal refactoring: new subdirectory under `scripts/codex-imagegen/` for modular CLI glue (cache, logger, parser, validator).
- Minor improvements to configuration, documentation, and extensibility.
v2.0.0
v2.0.0: rename baoyu-imagine→baoyu-image-gen and baoyu-image-cards→baoyu-xhs-images; cross-skill image-backend examples updated
v1.56.4
**[Major update, multiple new providers and batch mode]**
- Added support for new providers: Azure OpenAI, OpenRouter, Z.AI GLM-Image, MiniMax, Jimeng (即梦), Seedream (豆包).
- Introduced batch mode: generate multiple images using a batch JSON file (`--batchfile`), with adjustable concurrency (`--jobs`).
- Expanded configuration and environment variable options for new providers and flexible per-provider model/endpoint overrides.
- Updated reference-image identity preservation guidelines and safeguards, with improved prompt handling for real subjects.
- Enhanced provider auto-selection, fallback, and error handling logic; clarified user input tool usage and setup sequence.
- SKILL.md updated with in-depth setup, usage, and migration notes; now marked as [Deprecated: use baoyu-imagine].
v1.0.0
Initial release of baoyu-image-gen: a multi-provider API-based image generation tool.
- Supports image generation via OpenAI, Google, DashScope, and Replicate APIs.
- Allows text prompts, reference images, aspect ratios, image size, and quality presets.
- Provider, model, and defaults configurable via CLI flags, EXTEND.md, and environment variables.
- Ensures a blocking setup with EXTEND.md before image generation; guides first-time configuration.
- Sequential generation is default; parallel mode available upon explicit user request.
- Robust error handling with user setup hints and auto-retries for failed generations.
Metadata
Frequently Asked Questions
What is Baoyu Image Gen?
AI image generation with OpenAI GPT Image 2, Azure OpenAI, Google, OpenRouter, DashScope, Z.AI GLM-Image, MiniMax, Jimeng, Seedream, Replicate and Agnes APIs... It is an AI Agent Skill for Claude Code / OpenClaw, with 4168 downloads so far.
How do I install Baoyu Image Gen?
Run "/install baoyu-image-gen" in the OpenClaw or Claude Code chat to install it in one step — no extra setup required.
Is Baoyu Image Gen free?
Yes, Baoyu Image Gen is completely free, licensed under MIT-0. You can download, install and use it at no cost.
Which platforms does Baoyu Image Gen support?
Baoyu Image Gen is cross-platform and runs anywhere OpenClaw / Claude Code is available (cross-platform).
Who created Baoyu Image Gen?
It is built and maintained by Jim Liu 宝玉 (@jimliu); the current version is v2.1.0.
More Skills