DeepSeek Harness plugin

dsh-image-describe

DeepSeek Harness 宿主插件:让不支持图片输入的纯文本主模型也能"看图"(describe_image 工具 + 图片标记替换)

Jump to install

Source facts

Repository
sd1g1/dsh-image-describe
Latest update
Aug 21, 2026
Category
Vision & Multimodal
GitHub stars
1
Format
plugin
Catalog evidence
Upstream dsh.bundle evidence
Evidence path
package.json#dsh.bundle
Checked against
0.1.0-rc.8
Upstream check date
2026-08-20

This evidence comes from the upstream catalog. This site has not installed, run, or security-reviewed the plugin.

Install

Start with a prompt that asks an agent to review the GitHub repository and source. Switch to the command if you want to install it yourself.

Copy this prompt into DSH, Codex, or another agent and ask it to review the GitHub repository and source first.

Do not install or run any commands yet. Read this plugin's GitHub repository, README, and relevant source code. Then answer the questions below clearly and directly so I can decide whether it fits my needs:

1. What is this plugin, and what problem does it solve?
2. Who is it for, and what are its typical use cases?
3. How is it used after installation? Include one minimal example.
4. What known limitations or privacy, security, compatibility, or maintenance risks does it have?
5. Give a clear recommendation: recommend, conditionally recommend, or do not recommend, with reasons.

Distinguish statements documented by the repository, inferences from source code, and unknowns. If evidence is insufficient, say so explicitly. Do not guess or simply repeat the README.

GitHub: https://github.com/sd1g1/dsh-image-describe
Plugin: dsh-image-describe
Author: sd1g1

Check the source files

Read the README and other files from this plugin directory before installing.

File explorer3 files
README.mdSource · read only

dsh-image-describe

DeepSeek Harness 宿主插件:让不支持图片输入(纯文本)的主模型也能通过 describe_image 工具“看图”。

安装

npx @deepseek-ai/dsh plugin --profile web add github:sd1g1/dsh-image-describe

配置

首次启动会自动创建 ~/.dsh/image-describe.json(路径遵循 $DSH_HOME)。模板把 vision.provider / vision.model 留空,此时插件休眠,不影响正常请求;填写视觉模型后保存即热生效,无需重启:

{
  "vision": { "provider": "my-vision-provider", "model": "vision-model" },
  "prompt": "请详细描述这张图片的内容",
  "maxTokens": 1500,
  "image": { "autoResize": true }
}
  • vision.provider / vision.model:必填;留空 = 休眠
  • prompt:可选,默认兜底指令
  • maxTokens:可选,默认 1500
  • image.autoResize:可选,默认 true。发送给视觉模型前沿用 pi-coding-agent 的图片处理方法:大图先缩到 2000x2000,再在 PNG/多档 JPEG 中选最小编码;处理失败时回退原图,不影响看图。设为 false 关闭

调用方式

describe_imageattachmentpath 二选一:

  • attachment:使用之前标记里出现过的 attachmentId,无需用户重新发送图片。插件会依次搜索当前会话的派生消息、完整事件日志(被压缩的历史仍可解析)以及当前已加载的其他会话。
  • path:直接描述本地图片文件(绝对路径,或相对会话 cwd 的相对路径;支持 png/jpeg/webp/gif)。插件通过宿主 fs 服务读取文件,再交给附件服务校验并转为视觉模型可读的引用;与 attachment 一样,描述前会按 image.autoResize 预处理。

验证

  • 给纯文本模型发送一张图片:请求会被替换为 attachmentId 标记,模型可调用 describe_image 获取图片描述;之后让模型直接凭该 attachmentId 再次调用,也应成功。
  • 让模型调用 describe_image({ path: "/绝对路径/图片.png" }),应返回该图片的描述。

卸载

npx @deepseek-ai/dsh plugin --profile web remove @local/dsh-image-describe