zabr1314

小红书图文卡片自动生成器。将文章/观点/数据转换成高质量小红书图文笔记（1080×1440，3:4比例）。 使用场景：(1) 将长文/阅读笔记改写成小红书风格并生成配图 (2) 根据热点话题从零创作小红书内容 (3) 批量生成系列图文卡片。 触发词：小红书、xhs、图文卡片、生成配图、小红书内容、自媒体内容

Design and orchestrate multi-agent systems for complex tasks. Use when (1) designing agent dispatch strategies (spawn/fork/direct), (2) building agent communication protocols, (3) implementing agent lifecycle management (create/execute/complete/cleanup), (4) orchestrating verification and adversarial agent patterns, (5) building team-based agent systems. Triggers on phrases like multi-agent, agent orchestration, spawn agent, fork agent, agent communication, team agents, agent lifecycle, verification agent.

Build AI agent execution engines that turn prompts into behavior. Use when (1) designing the tool-use loop for an agent system, (2) implementing prompt caching strategies for cost optimization, (3) building the prompt assembly pipeline (system prompt + tool prompts + memory + dynamic context), (4) debugging why an agent's behavior doesn't match its prompt, (5) optimizing agent latency and token usage. Triggers on phrases like agent execution engine, tool-use loop, prompt caching, prompt assembly, streaming API, context management, agent architecture.

Design secure tool systems for AI agents using the dual-defense pattern (prompt guidance + code enforcement). Use when (1) designing tool prompts with preference chains and safety constraints, (2) implementing code-layer security checks for agent tool calls, (3) building tool permission and sandbox systems, (4) auditing existing tool security architectures. Triggers on phrases like tool security, agent tool design, preference chain, sandbox, tool permission, dual defense.

Design and write high-quality prompts for AI agents and coding assistants. Use when (1) creating new system prompts, tool prompts, agent prompts, skill prompts, or memory prompts, (2) reviewing or auditing existing prompts for quality, (3) debugging prompt behavior issues like model ignoring instructions or gold-plating, (4) migrating prompts between frameworks. Triggers on phrases like write a prompt, improve this prompt, prompt engineering, system prompt, agent prompt, prompt audit.

Evaluate and benchmark AI agent prompts with scientific rigor. Use when (1) assessing whether a prompt achieves its intended behavior, (2) comparing prompt versions with controlled experiments, (3) building an evaluation suite for prompt quality, (4) auditing existing prompts for weaknesses. Triggers on phrases like evaluate prompt, prompt benchmark, prompt audit, test prompt quality, compare prompts, prompt experiment.

个人知识引擎。搜索已有概念、添加新概念（自动去重）、查看信念状态、执行知识蒸馏、生成可视化图谱。当你需要：(1) 查找之前积累的知识和洞察 (2) 记录新的学习和思考 (3) 追踪信念的变化 (4) 从碎片概念中提炼高层规律 (5) 可视化知识图谱