dsh-opencode-eyes

DeepSeek Harness plugin: gives a text-only model eyes. It mounts the bundled opencode-eyes MCP server, which sends an image file to StepFun's step-3.7-flash vision model and returns a detailed description, as the tool mcp__opencode_eyes__describe_image. 适合需要让纯文本模型通过工具调用来识别和描述图片的用户。

Package

dsh-opencode-eyes

Compatibility

Unverified

Harness peer range

>=0.1.5-rc.1 <0.2.0-0 || >=0.2.0-rc.0 <0.3.0-0

Cordis peer range

>=4.0.2

Version

0.1.0

License

MIT

Last updated

Oct 8, 2026

Configuration

Key	Environment variable	Default	Meaning
`python`	—	discovered	interpreter that runs the server
`apiKey`	`STEP_API_KEY`	(empty)	StepFun credential; required by `describe_image`
`model`	`STEP_MODEL`	`step-3.7-flash`	model id sent in the request
`timeoutSeconds`	`STEP_TIMEOUT`	`120`	the server's own HTTP timeout
`maxDimension`	`STEP_MAX_DIMENSION`	`2048`	images are downscaled to this edge length
`jpegQuality`	`STEP_JPEG_QUALITY`	`85`	JPEG quality of the re-encoded image
`toolCallTimeoutMs`	—	`300000`	DSH's per-call budget; keep it above `timeoutSeconds`
`env`	—	`{}`	raw environment passthrough for anything else

Every field is optional and lives in the loader row. For example, in cordis.patch.yml:

- id: dsh-opencode-eyes
  name: 'dsh-opencode-eyes'
  config:
    apiKey: 'sk-...'
    toolCallTimeoutMs: 300000

dsh-opencode-eyes

Install

README

Configuration