lijiaqi-AI-Product-Manager/dsh-agent-compare

C3

同一需求,两版真跑,体验后审查采用。Two live builds for one task in DeepSeek Harness.

★ 100+ · lijiaqi-AI-Product-Manager/dsh-agent-compare source on GitHub · this plugin in the registry

lijiaqi-AI-Product-Manager/dsh-agent-compare is a DeepSeek Harness plugin rated C3 — powerful capability combined with sensitive behavior. It patches the dsh runtime, can rewrite the system prompt, executes system commands, starts a network server.

Installable plugin — declares a dsh.bundle manifest

What it can do

CapabilityFlagEvidence
patches the dsh runtimeruntime_patch./cordis.patch.yml
can rewrite the system promptprompt_surfacesystem-prompt/assemble
executes system commandsexec×9 in authored code, e.g. src/core/preview.ts:6, src/core/proc.ts:1
starts a network servernet_server×2 in authored code, e.g. src/core/preview.ts:16, src/core/preview.ts:19
reads credential-class env vars in development tooling the package does not shiptoken_env_toolingin development tooling the package does not ship: AGENT_COMPARE_TEST_TOKEN

Tool registrations

4 static tool registration point(s).

Services it injects

agents remote slots tools

Hooks it attaches

agent/pre-step system-prompt/assemble

Environment variables it reads

SystemRoot AGENT_COMPARE_WINDOWS_PIPE_COMPAT Path PATHEXT AGENT_COMPARE_PACKAGE_RUNTIME XDG_RUNTIME_DIR AGENT_COMPARE_HOME ProgramFiles ProgramData AGENT_COMPARE_WIN32_PROCESS_MODULE

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.