← 返回 Skills 市场
125
总下载
0
收藏
0
当前安装
5
版本数
在 OpenClaw 中安装
/install oatda-vision-analysis
功能描述
Analyze images using vision-capable AI models through OATDA's unified API. Triggers when the user wants to analyze, describe, or understand images; extract t...
安全使用建议
This skill is internally consistent, but review the following before installing: 1) Ensure you trust OATDA and are comfortable giving the skill an OATDA API key (it will be used in outbound HTTP calls). 2) Confirm ~/.oatda/credentials.json only contains the credentials you intend the skill to access. 3) Be aware the skill will perform network requests to https://oatda.com; avoid sending sensitive images you would not want transmitted. 4) Keep agent autonomy settings in mind — although this skill is benign and not always-included, an agent invoking it can send images automatically to the external API. Rotate/revoke the API key if you later uninstall or stop trusting the integration.
功能分析
Type: OpenClaw Skill
Name: oatda-vision-analysis
Version: 1.0.6
The skill bundle for 'oatda-vision-analysis' is a legitimate tool for interacting with the OATDA vision API. It includes security best practices such as validating image URLs to prevent SSRF (rejecting local/internal IPs) and explicitly declaring its need for credentials in ~/.oatda/credentials.json. No indicators of malicious intent, data exfiltration to unauthorized endpoints, or obfuscation were found in SKILL.md or _meta.json.
能力评估
Purpose & Capability
Name/description (vision analysis via OATDA) matches the declared requirements: curl and jq for making/parsing HTTP calls, an OATDA_API_KEY, and a ~/.oatda/credentials.json file. These are expected for a small CLI-style wrapper that reads a local credentials file and calls an external API.
Instruction Scope
SKILL.md instructions are narrowly scoped: they show how to resolve the API key (from env or the declared credentials.json), validate image URLs, assemble the JSON body, and call the documented /api/v1/llm/image endpoint. The only file and env accessed are the ones declared. It does instruct runtime curl calls to an external endpoint (expected for this purpose).
Install Mechanism
This is an instruction-only skill with no install spec or downloads. That is the lowest-risk pattern and is proportionate to the functionality.
Credentials
Only OATDA_API_KEY and a single config path (~/.oatda/credentials.json) are required, which directly map to authenticating to the OATDA API. No unrelated secrets, cloud credentials, or other service tokens are requested.
Persistence & Privilege
always is false and the skill does not request persistent/system-wide changes or access to other skills' configs. It runs ad-hoc curl commands and reads a declared local credentials file — expected for this type of skill.
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 oatda-vision-analysis - 安装完成后，直接呼叫该 Skill 的名称或使用
/oatda-vision-analysis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
v1.0.6
Re-publish with config path declaration for trust scan refresh
v1.0.3
Declare credentials.json file access in manifest
v1.0.2
Declare OATDA_API_KEY credential in metadata for trust scan
v1.0.1
Restore concrete model IDs with fallback note
v1.0.0
Initial release
元数据
常见问题
OATDA Vision Analysis 是什么？
Analyze images using vision-capable AI models through OATDA's unified API. Triggers when the user wants to analyze, describe, or understand images; extract t... 它是一个面向 Claude Code / OpenClaw 的 AI Agent Skill 插件，目前累计下载 125 次。
如何安装 OATDA Vision Analysis？
在 OpenClaw 或 Claude Code 对话框中运行命令「/install oatda-vision-analysis」即可一键安装，无需额外配置。
OATDA Vision Analysis 是免费的吗？
是的，OATDA Vision Analysis 完全免费，采用 MIT-0 许可证，可自由下载、安装和使用。
OATDA Vision Analysis 支持哪些平台？
OATDA Vision Analysis 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 OATDA Vision Analysis？
由 devcsde（@devcsde）开发并维护，当前版本 v1.0.6。
推荐 Skills