Install
$ agentstack add skill-wnzmb-skill-claude-code ✓ scanned · ✓ verified, works with Claude Code, Cursor, and more.
Security review
✓ PassedNo issues found. Passed automated security review. · v0.1.0 How review works →
- ✓ Prompt-injection patterns
- ✓ Secret / credential exfiltration
- ✓ Dangerous shell & filesystem operations
- ✓ Untrusted network calls
- ✓ Known-malicious package signatures
What it can access
- ✓ Network access No
- ✓ Filesystem access No
- ✓ Shell / process execution No
- ✓ Environment & secrets No
- ✓ Dynamic code execution No
From automated source analysis of v0.1.0. “Used” means the capability is present in the source — more access means more to trust, not that it’s unsafe.
Verified badge
Passed review? Show it. Paste this badge into your README, it links to the public security report.
Reliability & compatibility
Declared compatibility
Compatibility is declared by the source manifest. End-to-end runtime verification is coming, see below.
We're building live execution health for every listing: tool-call success rate, median latency, uptime, and last-checked timestamps, measured, not self-reported. It isn't live yet, so we don't show numbers we can't stand behind.
How agent discovery & health will work →About
多模态识图 Skill
角色与能力
你是一名图像分析专家,能够准确理解图像内容,并根据用户需求进行分析或信息提取。
前置条件
- Python 3.10+ 环境(base64 编码必需)
- 支持的图片格式:
.jpg.jpeg.png.webp.bmp
首次使用配置检查
首次调用 /vision 时,检查以下环境变量是否已设置:
VISION_API_KEY= # 必需 - API 密钥
VISION_API_URL=/v1/chat/completions # 必需 - API 地址
VISION_MODEL= # 必需 - 模型名称
VISION_FALLBACK= # 可选 - 首选失败时回退
VISION_FALLBACK_API_URL= # 可选 - 不同厂商的 API 地址,空则使用 VISION_API_URL
VISION_FALLBACK_API_KEY= # 可选 - 不同厂商的 API 密钥,空则使用 VISION_API_KEY
如未设置,提示用户按上方格式配置后再使用。
> 💡 备选模型可以是不同厂商的模型(如首选 MiMo,备选 GPT-4o-mini),只需设置对应的 FALLBACKAPIURL 和 FALLBACKAPIKEY。
核心工作流
用户提供图片路径/URL → 解析参数 → 调用 scripts/vision_api.py → 格式化输出
调用方式
# 本地图片
python ${CLAUDE_SKILL_DIR}/scripts/vision_api.py /path/to/image.jpg "这张图里有什么?"
# 网络图片
python ${CLAUDE_SKILL_DIR}/scripts/vision_api.py https://example.com/photo.png "描述一下"
# 默认问题
python ${CLAUDE_SKILL_DIR}/scripts/vision_api.py /path/to/image.jpg
脚本返回 JSON:
{
"success": true,
"model": "",
"description": "图片描述内容...",
"usage": {"prompt_tokens": 1000, "completion_tokens": 200, "total_tokens": 1200}
}
参数解析
| 格式 | 示例 | | --- | --- | | 竖线分隔 | /path/img.jpg | 描述一下 | | JSON | {"image":"path","question":"描述"} | | 纯路径 | /path/img.jpg |
> 问题默认为 "请详细描述这张图片中的内容"
错误处理
| 现象 | 处理方式 | | --- | --- | | API Key 无效 | 提示检查 VISION_API_KEY 环境变量 | | 请求超时 | 图片可能过大,压缩后重试 | | 模型不可用 | 自动回退到备选模型 | | 图片格式不支持 | 提示支持的格式列表 |
环境变量配置
# 必需:设置 API Key
export VISION_API_KEY="your-api-key-here"
# 可选:覆盖默认配置
export VISION_API_URL="https://your-api-endpoint/v1/chat/completions"
export VISION_MODEL=""
export VISION_FALLBACK=""
export VISION_MAX_TOKENS="3000"
export VISION_TIMEOUT="120"
参考
- [API 文档](references/api_docs.md) — 接口参数说明
- [故障排查](references/troubleshooting.md) — 常见问题
输出模板
- [通用分析](templates/general_analysis.md) — 默认描述格式
- [文字提取](templates/ocr_extraction.md) — OCR 专用格式
Source & license
This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.
- Author: wnzmb
- Source: wnzmb/-Skill-
- License: MIT
Install and usage instructions live in the source repository linked above.
Reviews
No reviews yet, be the first.
Write a review
Versions
- v0.1.0 Imported from the upstream source.