Schema | vision-mcp

vision-mcp

Describes the environment variables required to run the server.

Name	Required	Description	Default
`VISION_MODEL`	No	Model identifier for the vision backend (default depends on profile)
`Z_AI_API_KEY`	No	Alternative API key (alias for VISION_API_KEY)
`VISION_API_KEY`	No	API key for the vision backend (alias Z_AI_API_KEY)
`VISION_PROFILE`	No	Backend profile: glm, kimi, mimo, openai, generic	glm
`VISION_BASE_URL`	No	Base URL for the OpenAI-compatible endpoint (default depends on profile)
`VISION_MAX_EDGE_PX`	No	Overview downsample longest edge in pixels	1568
`VISION_ALLOWED_DIRS`	No	Directories allowed for local image paths (semicolon or colon separated)	current working directory
`VISION_MAX_IMAGE_MB`	No	Maximum image file size in MB	10
`VISION_MAX_VIDEO_MB`	No	Maximum video file size in MB	50
`VISION_VIDEO_FRAMES`	No	Frames sampled per video	8
`VISION_MAX_ZOOM_ROUNDS`	No	Maximum agentic zoom rounds	3
`VISION_ALLOW_URL_PASSTHROUGH`	No	Forward image URLs to the backend (default: download + SSRF-check → data URI)	false

Guidance the server publishes about itself, which clients place ahead of the tool catalog so the model reads it before choosing anything.

This server publishes no instructions, or was last inspected before Glama recorded them.

Features and capabilities supported by this server

Protocol revision2025-11-25

Capability	Details
`tools`	{ "listChanged": true }

Functions exposed to the LLM to take actions

Name	Description
ui_to_artifactA	把一张 UI 截图/设计稿转成可运行代码或结构化规格。当宿主拿到界面图、需要据此生成或还原前端实现时使用。
extract_text_from_screenshotA	逐字提取截图中的文本（代码、终端、报错、文档等），保留阅读顺序与布局。需要把图里的文字读出来时使用。
diagnose_error_screenshotA	分析报错/异常截图，给出根因、逐字错误原文、位置和可执行的修复步骤。处理崩溃/红屏/堆栈截图时使用。
understand_technical_diagramB	解读架构图/流程图/UML/ER/时序图等：节点、连线、流程与设计意图。需要读懂一张技术示意图时使用。
analyze_data_visualizationA	读懂图表/仪表盘并抽取数据与洞察（趋势、异常、数值）。需要从图表里读出数字或结论时使用。
ui_diff_checkA	对比两张 UI 截图（A 基准 / B 对照），逐条列出视觉与实现差异及可能的回归。做视觉回归/前后对比时使用。
image_analysisA	通用兜底：理解任意图片并回答问题。不确定用哪个专用工具，或只是想问一张图时使用。
video_analysisA	理解一段视频（时序+画面）并回答问题。无原生视频能力的后端会自动走帧采样。需要分析录屏/短视频时使用。

Interactive templates invoked by user choice

Name	Description
No prompts

Contextual data attached and managed by the client

Name	Description
No resources

We provide all the information about MCP servers via our MCP API.

curl -X GET 'https://glama.ai/api/mcp/v1/servers/Pelican0126/vision-mcp'

If you have feedback or need assistance with the MCP directory API, please join our Discord server