omdsh-dev/dsh-custom-tool
C3Create and manage sandboxed JavaScript tools for DeepSeek Harness with a Monaco editor and model-driven tool lifecycle.
★ 10+ · omdsh-dev/dsh-custom-tool source on GitHub · this plugin in the registry
omdsh-dev/dsh-custom-tool is a DeepSeek Harness plugin rated C3 — powerful capability combined with sensitive behavior. It patches the dsh runtime, can rewrite the system prompt, executes commands in development tooling the package does not ship, uses eval / new Function.
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 |
| executes commands in development tooling the package does not ship | exec_tooling | ×3 in development tooling the package does not ship, e.g. scripts/build.mjs:12, scripts/build.mjs:12 |
| uses eval / new Function | eval | ×1 in authored code, e.g. src/client/validate.ts:35 |
| base64 decoding in build output only | base64_decode_bundled | ×2 in build output only, e.g. lib/index.js:395, lib/index.js:396 |
Services it injects
invariants locale settings settingsScope slots systemPrompt tools
Hooks it attaches
agent/created agent/disposed theme/change
Environment variables it reads
DSH_HOME VSCODE_CWD
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.