ccch1mneyyy/dsh-TUI
C3DSH 官方公众号收录的 TUI 补位插件：Claude Code 风，鲸鱼顶栏/实时状态/流式思考/双击 Esc 回滚/上下文进度+TPS。npm 一键装。 DSH official WeChat featured TUI plugin — Claude Code style: whale bar, live status, streaming thoughts, double-Esc rol
★ 1.9k · ccch1mneyyy/dsh-TUI source on GitHub · this plugin in the registry
ccch1mneyyy/dsh-TUI is a DeepSeek Harness plugin rated C3 — powerful capability combined with sensitive behavior. It patches the dsh runtime, can rewrite the system prompt, can spawn subprocesses, executes system commands.
What it can do
| Capability | Flag | Evidence |
|---|---|---|
| patches the dsh runtime | runtime_patch | ./cordis.patch.yml |
| can rewrite the system prompt | prompt_surface | subprocess, system-prompt/assemble |
| can spawn subprocesses | subprocess_service | inject: subprocess |
| executes system commands | exec | ×55 src, e.g. bin/dsh-tui.js:28, bin/dsh-tui.js:28 |
| uses eval / new Function | eval | ×1 src, e.g. scripts/verify-cordis-approval.mjs:44 |
| decodes base64 payloads | base64_decode | ×4 src, e.g. scripts/verify-copy-on-select.mjs:137, scripts/verify-copy-on-select.mjs:138 |
| reads credential-class env vars | token_env | DEEPSEEK_API_KEY |
Services it injects
agents invariants skills subprocess tools
Hooks it attaches
agent/disposed agent/pre-step agent/request agent/status approval/request commands/change internal/listener llm/stream session/event skills/change system-prompt/assemble
Outbound domains
css-tricks.com mirror.example.com env-registry.example.com lower-registry.example.com sw.kovidgoyal.net anthropic.slack.com api.example.com mitchellh.com a.com b.com
Environment variables it reads
FORCE_COLOR USERPROFILE DSH_TUI_LANG TERM_PROGRAM WAYLAND_DISPLAY DISPLAY DSH_HOME EDITOR DSH_TUI_THEME DEEPSEEK_API_KEY
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.