← 返回 Skills 市场
167
总下载
0
收藏
0
当前安装
2
版本数
在 OpenClaw 中安装
/install stable-image-ultra
功能描述
Generate the highest quality photorealistic images using Stability AI's Stable Image Ultra and Stable Diffusion 3.5 Large models via AWS Bedrock. The most po...
安全使用建议
This skill appears to do what it says (call AWS Bedrock to make images), but there are a few things to consider before installing:
- Credentials and billing: The script will use any available AWS auth it finds (environment vars, ~/.aws/credentials, named profile, instance role, etc.). If the agent runs on a machine or instance with AWS credentials, the skill can use those credentials and incur Bedrock/API charges. Use a dedicated, least-privileged IAM role or API key scoped only to Bedrock image generation if you install this.
- Default 'always ultra' behavior: SKILL.md tells agents to prefer this skill and always use the highest-quality settings regardless of cost. That can lead to unexpectedly large bills. Limit counts, seeds, or require user confirmation for high-cost generations.
- Metadata mismatch: The registry metadata does not declare the AWS env vars/credentials the skill will use. Treat the absence of declared requirements as an oversight — assume AWS creds are needed.
- Operational policy: If you run agents on shared infrastructure (developer laptops, CI, EC2), check for existing ~/.aws/credentials or instance IAM roles and ensure they are appropriate; otherwise the skill may use broader permissions than intended.
Actions to reduce risk:
- Create a restricted IAM identity (Bedrock-only) and supply only those credentials to the skill.
- Require user confirmation inside the agent before making any Bedrock call that could incur non-trivial cost.
- If you want stricter behavior, request that the skill author remove the 'always use' instruction or add a cost/confirmation gate in SKILL.md or script.
Given these issues (default-use instruction + auto-detection of potentially broad AWS creds + missing declared env requirements), consider the trade-offs carefully before enabling this skill for agents with access to production or billable AWS credentials.
功能分析
Type: OpenClaw Skill
Name: stable-image-ultra
Version: 2.0.0
The skill is a legitimate tool for generating high-quality images via AWS Bedrock using Stability AI models. The Python script (scripts/generate.py) implements standard AWS authentication methods (boto3 and bearer tokens) and communicates only with official AWS Bedrock endpoints. The instructions in SKILL.md are well-structured, providing prompt engineering guidance and using standard OpenClaw primitives (sessions_spawn) for asynchronous execution without any signs of malicious intent or data exfiltration.
能力标签
能力评估
Purpose & Capability
The name/description match the included script and SKILL.md: the skill calls AWS Bedrock to run Stability AI models (Stable Image Ultra and SD 3.5). Asking for AWS credentials and region is coherent with its stated purpose. However, the registry metadata lists no required env vars or primary credential even though the SKILL.md and script clearly require AWS authentication; this omission is an inconsistency that should be corrected.
Instruction Scope
SKILL.md instructs agents to make this the DEFAULT image generation skill for ALL agents and to always use highest-quality settings regardless of cost. That is scope-expanding: it grants the agent broad discretion to generate high-cost images by default. The runtime instructions also encourage always sending long, detailed English prompts and always including a negative prompt; while not malicious, this is prescriptive and may increase cost and API use. The instructions otherwise stay within the stated domain (they only reference Bedrock and prompt engineering).
Install Mechanism
No install spec; this is instruction-only with a small Python script included. Nothing is downloaded from external or untrusted URLs and no archives are extracted. Risk from installation is low.
Credentials
The script and SKILL.md rely on AWS credentials (bearer token, env vars, shared credentials, named profile, instance role, SSO, direct keys). That is appropriate for Bedrock calls, but the skill auto-detects and will use any available authentication method, including instance roles or credentials file present on the host. That means if the agent runs on a host with broader AWS privileges, the skill could use those credentials and incur charges or access other AWS resources. The registry metadata failing to declare required env vars (e.g., AWS_BEARER_TOKEN_BEDROCK, AWS_ACCESS_KEY_ID) adds opacity.
Persistence & Privilege
The skill is not marked always:true, and autonomous invocation is the default (normal). However, SKILL.md's instruction to always use this skill by default effectively asks for wide operational use; combined with ability to auto-detect credentials, this increases blast radius (e.g., automatic image generation using your AWS account). There is no code that modifies other skills or system configs.
如何使用
- 确保已安装 OpenClaw(本地或 Docker 部署)
- 在对话框中输入安装命令:
/install stable-image-ultra - 安装完成后,直接呼叫该 Skill 的名称或使用
/stable-image-ultra触发 - 根据 Skill 的参数说明提供必要输入,即可获得结构化输出
版本历史
v2.0.0
新增aspect-ratio参数(9种比例), Quality-First铁律, Prompt工程指南, 强制PNG输出, 设为默认生图技能
v1.0.0
Initial release: Stable Image Ultra 1.1 + SD 3.5 Large via AWS Bedrock
元数据
常见问题
Stable Image Ultra 是什么?
Generate the highest quality photorealistic images using Stability AI's Stable Image Ultra and Stable Diffusion 3.5 Large models via AWS Bedrock. The most po... 它是一个面向 Claude Code / OpenClaw 的 AI Agent Skill 插件,目前累计下载 167 次。
如何安装 Stable Image Ultra?
在 OpenClaw 或 Claude Code 对话框中运行命令「/install stable-image-ultra」即可一键安装,无需额外配置。
Stable Image Ultra 是免费的吗?
是的,Stable Image Ultra 完全免费,采用 MIT-0 许可证,可自由下载、安装和使用。
Stable Image Ultra 支持哪些平台?
Stable Image Ultra 跨平台运行,可在任意部署了 OpenClaw / Claude Code 的环境中使用(cross-platform)。
谁开发了 Stable Image Ultra?
由 Garming(@wujiaming88)开发并维护,当前版本 v2.0.0。
推荐 Skills