loadingvx/deepseek-harness-workbench-plugin
C3Deepseek-harness-workbench-plugin
★ 10+ · loadingvx/deepseek-harness-workbench-plugin source on GitHub · this plugin in the registry
loadingvx/deepseek-harness-workbench-plugin is a DeepSeek Harness plugin rated C3 — powerful capability combined with sensitive behavior. It patches the dsh runtime, can rewrite the system prompt, gates tool execution, executes system commands.
Installable plugin — declares a dsh.bundle manifest
What it can do
| Capability | Flag | Evidence |
|---|---|---|
| patches the dsh runtime | runtime_patch | ./cordis.patch.yml |
| can rewrite the system prompt | prompt_surface | systemPrompt, webServer, tools/pre-execute |
| gates tool execution | tool_gate | hook: tools/pre-execute |
| executes system commands | exec | ×10 in authored code, e.g. src/host/external-open.ts:1, src/host/external-open.ts:89 |
| uses eval / new Function | eval | ×1 in authored code, e.g. src/shared/canvas-prepare.ts:61 |
Services it injects
agentDefaultModel agents commands inputTriggers llm locale remote sessions slots systemPrompt tools webServer
Hooks it attaches
agent/created agent/disposed agent/pre-step agent/request command/executed tools/pre-execute tools/result
Outbound domains
api.deepseek.com www.npmjs.com platform.deepseek.com
Environment variables it reads
WSL_DISTRO_NAME SYSTEMROOT
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.