lcestou/dsh-oh-my-claude

C3

Claude Code for dsh (DeepSeek Harness): runs your logged-in Claude Code CLI as a model provider, with live models, session resume, approvals, rewind, memory and remote SSH workspaces

★ 10+ · lcestou/dsh-oh-my-claude source on GitHub · this plugin in the registry

lcestou/dsh-oh-my-claude is a DeepSeek Harness plugin rated C3 — powerful capability combined with sensitive behavior. It patches the dsh runtime, executes system commands, starts a network server, reads credential-class env vars.

Installable plugin — declares a dsh.bundle manifest

What it can do

CapabilityFlagEvidence
patches the dsh runtimeruntime_patch./cordis.patch.yml
executes system commandsexec×16 in authored code, e.g. src/dsh.ts:196, src/keeper.ts:16
starts a network servernet_server×2 in authored code, e.g. src/keeper.ts:224, src/mcp.ts:409
reads credential-class env varstoken_envANTHROPIC_API_KEY

Tool registrations

0 confirmed registration point(s); analysis is incomplete.

Services it injects

agents approval attachments commands llm modelDirectories sessions slots userQuestions workspaces

Outbound domains

api.anthropic.com dsh.other-box.lan claude.ai code.claude.com status.anthropic.com downloads.claude.ai api.github.com api.githubcopilot.com mcp.notion.com mcp.sentry.dev

Environment variables it reads

DSH_OMC_STATE_DIR CLAUDE_CONFIG_DIR ANTHROPIC_API_KEY ANTHROPIC_BASE_URL XDG_RUNTIME_DIR CLAUDE_CODE_DISABLE_CLAUDE_MDS

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.