chenxiachan/thoughtdag
C3Your thinking deserves a map: an infinite canvas where LLM conversations grow into an editable thought graph. Wires are the context.
★ 500+ · chenxiachan/thoughtdag source on GitHub · this plugin in the registry
chenxiachan/thoughtdag 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, decodes base64 payloads.
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 | approval, systemPrompt, webServer |
| executes system commands | exec | ×58 in authored code, e.g. server.mjs:21, server.mjs:21 |
| decodes base64 payloads | base64_decode | ×3 in authored code, e.g. server.mjs:564, runtime/agents/ops.cjs:36 |
| starts a network server | net_server | ×1 in authored code, e.g. desktop/main.js:49 |
| reads credential-class env vars | token_env | ANTHROPIC_API_KEY, ANYSEARCH_API_KEY, DASHSCOPE_API_KEY, DEEPSEEK_API_KEY, GOOGLE_API_KEY, GOOGLE_GENERATIVE_AI_API_KEY |
Services it injects
agents approval attachments commands llm sessionController sessions systemPrompt tools uiWorkspace web webServer
Hooks it attaches
agent/assistant-stream approval/request session/event
Outbound domains
openrouter.ai open.bigmodel.cn chenxiachan.github.io react.dev api.moonshot.cn api.anysearch.com export.arxiv.org api.semanticscholar.org api.z.ai z.ai
Environment variables it reads
DSH_HOME TD_AGENT_IDLE_MS OPENAI_API_KEY ANTHROPIC_API_KEY GOOGLE_API_KEY GOOGLE_GENERATIVE_AI_API_KEY DEEPSEEK_API_KEY MOONSHOT_API_KEY OPENROUTER_API_KEY ZHIPU_SEARCH_ENGINE
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.