Skip to main content
Glama

Server Configuration

Describes the environment variables required to run the server.

NameRequiredDescriptionDefault
CANVAS_PORTNo写死画布端口(一般不用,插件会自己找)。对应配置文件里的 `canvas_port` 项。
XUANSHANG_CONFIRM_MODENo确认方式:`auto`、`elicitation`(只用 AI 工具确认框)、`popup`(只用 Windows 窗口)、`code`(只用确认码)。默认 `auto`。对应配置文件里的 `confirm_mode` 项。
XUANSHANG_CANVAS_CONFIGNo指定配置文件位置(配置项默认在 C:\Users\你的用户名\AppData\Roaming\xuanshang-canvas\config.json)。
XUANSHANG_CERTIFY_WAIT_SECONDSNo上传后最多等认证多少秒,默认 50。对应配置文件里的 `certify_wait_seconds` 项。
XUANSHANG_POPUP_TIMEOUT_SECONDSNo确认框/确认窗口多少秒不点算取消,默认 45。对应配置文件里的 `popup_timeout_seconds` 项。

Instructions

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.

Capabilities

Features and capabilities supported by this server

Protocol revision2025-11-25

CapabilityDetails
tools
{
  "listChanged": false
}
prompts
{
  "listChanged": false
}
resources
{
  "subscribe": false,
  "listChanged": false
}

Tools

Functions exposed to the LLM to take actions

NameDescription
canvas_statusA

查看画布是否在运行、是哪个程序(无限画布/内部版)、端口、版本是否兼容、项目数量。开始操作前先调一次。

list_projectsA

列出画布里的项目(名称、ID、节点数)。

get_graph_summaryB

项目节点摘要:每个视频节点的名字、参数、按顺序列出的接入素材(图片1、视频1……)、最近任务状态、 是否已有生成好的视频(可以当参考);以及认证图片节点、认证视频节点的名字和 ID。

list_certified_groupsC

项目的认证分组,以及组内每个素材(图片、视频)的认证状态(向玄上实时查)、是否已经放到画布上。

get_taskA

查某个任务,或某个节点最近一次任务的状态、进度和输出文件。只在用户问进度时调用,不要反复轮询。

create_projectA

在画布里新建一个空项目。已经有同名项目时会拒绝(避免以后选错项目),请换个名字或直接用原来的项目。

create_certified_groupA

在项目里新建一个认证分组(玄上那边也会同时建好),之后上传认证素材时选它。 注意:玄上没有删除分组的接口,建了就删不掉,名字要先和用户确认。同名分组已存在时会拒绝。

upload_certified_imageA

本地图片按原质量上传并认证(JPG/PNG/WEBP/BMP/GIF)。返回素材 ID、存储去向、上传核对结果、认证状态。 长边超过 6000px 的图会被画布压缩,插件会先请用户确认。上传后要再用 add_certified_asset_node 放到画布上。

upload_certified_videoA

本地参考视频按原文件上传并认证(不转码、不剪辑)。返回素材 ID、存储去向、大小核对结果、认证状态。 本地参考视频必须先认证才能接到视频节点上;画布上已生成的视频不用这一步。认证完再用 add_certified_asset_node 放到画布上。

add_certified_asset_nodeA

把已认证的素材(图片或视频)放到画布上,自动找空位,排成一列。只接受「已认证」的素材。

create_video_nodeA

新建 Seedance 视频生成节点并连好素材。会检查:引用格式(自动统一成画布的 (图片N) 写法)、引用编号和接入素材是否对得上、 素材是否已认证/已生成、参数是否合规;建完马上读回核对。不会提交生成。

submit_generationA

提交生成任务(花积分)。每次都要用户单独确认:插件会弹确认框/确认窗口,用户点确认才提交。 同一节点已有任务在跑时直接拒绝。提交后只返回任务 ID,不会自动盯进度。

create_final_videoA

把 Seedance 2.5 样片转成 1080p 成片(花积分),每次都要用户单独确认。画布会在样片节点旁边新建「成片生成」节点。

Prompts

Interactive templates invoked by user choice

NameDescription

No prompts

Resources

Contextual data attached and managed by the client

NameDescription

No resources

TDQS

A3.6/5.0

Scored across 13 tools

Disambiguation4/5

Each tool targets a fairly distinct resource and action: status vs. project listing vs. project detail, image vs. video upload, sample-video node vs. final-video node. The main potential confusion is among the read/status tools (canvas_status, list_projects, get_graph_summary, list_certified_groups), but their granularity differs enough to keep selection clear.

Naming Consistency4/5

Almost all tools follow a verb_noun snake_case pattern (create_project, list_projects, get_task, upload_certified_image, add_certified_asset_node, create_video_node, submit_generation). The one deviation is canvas_status, which uses a noun_status form instead of a verb, but it remains readable and unambiguous.

Tool Count5/5

13 tools sit squarely in the well-scoped range and each maps to a real step in the canvas workflow. No redundant or filler tools are present.

Completeness3/5

The core workflow is covered end to end: status, project/group creation, asset upload, node creation, submission, and final render. However, there are notable gaps: no delete/remove for projects, groups, nodes, or assets (the group deletion gap is explicitly acknowledged), and no bulk task listing or node-parameter update.

Maintenance

ActivityMaintained
ResponsivenessNo issues