JerryLiu369/agent-web-search
C2Agent-native multi-provider web search for AI agents (Claude Code, Hermes, DeepSeek Harness, OpenCode, Codex). Model grounding (Responses, ARK) + semantic search (Exa, Parallel) + keyless defaults via
★ 10+ · JerryLiu369/agent-web-search source on GitHub · this plugin in the registry
JerryLiu369/agent-web-search is a DeepSeek Harness plugin rated C2 — one powerful capability or sensitive behavior. It patches the dsh runtime, can rewrite the system prompt, command execution in build output only.
Installable plugin — declares a dsh.bundle manifest
What it can do
| Capability | Flag | Evidence |
|---|---|---|
| patches the dsh runtime | runtime_patch | ./integrations/dsh/cordis.patch.yml |
| can rewrite the system prompt | prompt_surface | systemPrompt |
| command execution in build output only | exec_bundled | ×2 in build output only, e.g. integrations/dsh/lib/bridge.js:1, integrations/dsh/lib/bridge.js:385 |
Tool registrations
1 confirmed registration point(s); analysis is incomplete.
integrations/dsh/lib/tool.js:246tool factoryintegrations/dsh/lib/tool.js:371Cannot establish the tools service binding
Services it injects
agents configForms credentials locale remote slots systemPrompt tools web
Hooks it attaches
agent/created agent/disposed
Outbound domains
open.bigmodel.cn mcp.exa.ai search.parallel.ai api.search.brave.com api.tavily.com api.perplexity.ai ydc-index.io generativelanguage.googleapis.com api.x.ai ark.cn-beijing.volces.com
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.