moton16/dsh-vision-mcp
C2Give your text-only LLM eyes: zero-dependency MCP server that reads images via external OpenAI-compatible vision models with automatic provider fallback, plus a one-command patch for DSH image-attachm
★ 1+ · moton16/dsh-vision-mcp source on GitHub · this plugin in the registry
moton16/dsh-vision-mcp is a DeepSeek Harness plugin rated C2 — one powerful capability or sensitive behavior. It decodes base64 payloads, reads credential-class env vars.
No dsh integration found — this repository carries the dsh-plugin topic but shows no sign of connecting to dsh
What it can do
| Capability | Flag | Evidence |
|---|---|---|
| decodes base64 payloads | base64_decode | ×2 in authored code, e.g. server.js:107, server.js:119 |
| reads credential-class env vars | token_env | VISION_API_KEY, VISION_MAX_TOKENS |
Outbound domains
api.openai.com
Environment variables it reads
DSH_CLI VISION_REASONING DSH_HOST_APIPROXY_INDEX VISION_API_KEY VISION_API_BASE VISION_MODEL VISION_MAX_TOKENS VISION_TEMPERATURE VISION_TIMEOUT_MS VISION_PROVIDERS
How to read this
Levels measure capability surface and transparency, not maliciousness. A C3 plugin can be entirely legitimate — a desktop shell genuinely needs subprocesses. The point is that you can see this before installing. See the levels explained and how dsh plugins work.
Findings come from static analysis of shipped code; nothing is executed. Think a flag is wrong? Open an issue — every flag cites the file and line it came from.