Research

在分享给干系人前对分析进行 QA 检查——方法论验证、准确性核实和偏差检测。适用于审查分析错误、检查幸存者偏差、验证聚合逻辑或准备可复现性文档。

基于飞书生态的**每周周报**自动生成 Skill：从当周**日程、任务、妙记、云文档**中提取工作内容，使用 `knowledge_answer` 兜底补全遗漏信息，最终调用 `doc_agent (writer)` 生成结构化的飞书云文档周报。

面向学术论文的**原文级调研（Deep Research）**能力模块。适用于：用户指定某篇论文要求"读透 + 追踪背景 + 扩展最新相关工作"，或用户给出某个学术 topic 要求系统性深度调研与写作综述。

Perform AI-powered web searches with real-time information using Perplexity models via LiteLLM and OpenRouter. This skill should be used when conducting web searches for current information, finding recent scientific literature, getting grounded answers with source citations, or accessing information beyond the model knowledge cutoff. Provides access to multiple Perplexity models including Sonar Pro, Sonar Pro Search (advanced agentic search), and Sonar Reasoning Pro through a single OpenRout...

Conduct comprehensive AI-powered research with citations via the Tavily CLI. Use this skill when the user wants deep research, a detailed report, a comparison, market analysis, literature review, or says "research", "investigate", "analyze in depth", "compare X vs Y", "what does the market look like for", or needs multi-source synthesis with explicit citations. Returns a structured report grounded in web sources. Takes 30-120 seconds. For quick fact-finding, use tavily-search instead.

Web search, content extraction, crawling, and deep research via the Tavily CLI. Use this skill whenever the user wants to search the web, find articles, research a topic, look something up online, extract content from a URL, grab text from a webpage, crawl documentation, download a site's pages, discover URLs on a domain, or conduct in-depth research with citations. Also use when they say "fetch this page", "pull the content from", "get the page at https://", "find me articles about", or refe...

Validate PromptSource-style prompt metadata and community quality constraints for natural-language prompts, metrics, answer choices, and target format.

Apply PromptSource-style templates across multiple examples and summarize browse, sourcing, and review diagnostics for prompt iteration.

Materialize PromptSource/P3-style template collections into prompted dataset records with template identity, metadata propagation, coverage counts, and rendering diagnostics.

Run a deterministic reduced LoRA training step that updates only low-rank factors and records loss evidence.

Assemble and validate a soft-mode reduced LoRA recovery with executable mechanism evidence and source-boundary logs.

Execute a bounded instruction-conditioned training proxy and ROUGE-L evaluation for BART0-style recovery.

Create task-level seen and unseen splits with leakage diagnostics for Natural Instructions style generalization experiments.

Evaluate Tent recovery artifacts for numeric metrics, source-boundary compliance, and mechanism-faithful proxy evidence.

Run a bounded CMA-ES proxy experiment that validates sampling, selection, CSA, and covariance adaptation evidence.

Run a deterministic reduced Gaussian-mixture recovery that exercises stochastic-interpolant construction, objectives, denoising, and sampling evidence.

Construct endpoint-valid stochastic interpolant samples with latent schedules and derivative targets for reduced or full generative recovery experiments.

Evaluate SR3-style proxy runs with consistency metrics, source-boundary records, and mechanism checks. This reusable skill supports bounded SR3 recovery experiments.

Build and validate SR3-style low-resolution conditioning paired with high-resolution targets for reduced or full recovery experiments.

Define continuous-time score-SDE schedules, perturbation kernels, reverse drifts, and probability-flow checks for recovery experiments.

Run a bounded synthetic MSFM proxy experiment that invokes generated coupling and Joint CFM loss skills.

Evaluate whether an MSFM recovery result exercised BatchOT coupling, Joint CFM loss, optimizer evidence, and soft-mode proxy constraints.

Run a bounded DDPM proxy experiment that exercises generated schedule, epsilon-loss, and reverse-step skills.

Run and validate a bounded GeDi proxy recovery that invokes generated GeDi modules and records mechanism-faithful evidence.