DeepSeek Harness plugin for local image / voice / music / SFX generation: the same character stays the same character across calls, degenerate output is refused rather than returned, and the GPU is untouched when idle. Hand the image half to any OpenAI-shaped API instead and its engine and 10.1 GiB of weights are never installed.
Audit citation sources in AI responses – extract URLs from the reply, score and rank them by threat level, and output a detailed report. Domains on your blocklist will be genuinely blocked – web tools such as web_search and web_fetch will directly refuse to access them. The blocklist is fully maintained by you.
Windows desktop UI context picker for AI agents: select windows, UI elements, or screen regions with hover highlighting, UI Automation, screenshots, local OCR, and user-approved MCP output. Works with OpenAI Codex, DeepSeek Harness, and other MCP-compatible clients.
Four terse levels plus independent reasoning-effort, web-search and subagent switches, on the conversation header and in the Settings page. No output ceiling.
Context window optimization for DeepSeek Harness agents. Sandboxes tool output in-process (only what you print enters the conversation), persists session memory in an FTS5 knowledge base, and enforces routing via a system-prompt routing section.
Automatically resumes replies cut off at the output token limit — built for self-hosted Ollama users with small output caps. · 回答被输出上限截断时自动续写，面向自部署 Ollama 用户。
