Research

Audit, organize, maintain, and optimize local knowledge bases, context folders, agent memory folders, project brains, research libraries, or source-of-truth file trees with zero knowledge loss. Use when asked to create a knowledge base from scratch, clean a messy source of truth, maintain Markdown knowledge files, add or route new source files, refresh indexes, reduce drift, restructure knowledge, streamline context engineering, reduce token consumption, improve agent startup speed, create no...

Comprehensive framework for conducting institutional-quality thematic investment research. Use when the user asks to: (1) Identify and validate investment themes, (2) Analyze thematic market opportunities and value chains, (3) Build a thematic stock universe and screen candidates, (4) Conduct company deep dives with thematic focus, (5) Value companies using theme-appropriate methodologies, (6) Assess theme-level and company-level risks, (7) Write thematic research reports, (8) Design portfoli...

Read and extract content from PDF investment reports — sell-side notes (GS/MS/JPM/UBS/Citi/BofA), 10-Ks, annual reports, fund factsheets, manager commentaries, and macro outlooks. Specialises in CHARTS, TABLES, and IMAGES that text-only extraction silently misses (most research charts are vector and invisible to pdfimages). Handles MULTI-REPORT TRIANGULATION when 2+ reports cover the same name, comparing forecasts, valuation methods, and surfacing coverage gaps. When the user has a company of...

Systematic framework for analyzing corporate networks to generate investment alpha. Use when analyzing supplier-customer relationships, director interlocks, executive networks, insider holdings patterns, or identifying M&A signals through board connections. Triggers on supply chain analysis, customer momentum, board interlock, director network, connected party analysis, relationship mapping, or network-based investment research.

Enter talk mode — conversation and research, no repo changes. ONLY invoke when the user explicitly types `/talk` or `$talk`; never auto-select from a natural-language question or design discussion.

Iterative measurement-driven improvement loop. Measure, profile, mutate, re-measure, commit. Works for performance, bundle size, complexity, test coverage — anything quantifiable. Use when the user wants to systematically improve a metric through repeated cycles of profiling and targeted changes.

Write essays, articles, and blog posts in Paul Graham's voice — plain, spoken prose that reasons out loud and lands surprising true claims, not AI filler. Use when asked to write or rewrite something "like Paul Graham", "in PG's style", or to make a piece read like a person thinking on the page rather than generated text. Built from a stylistic model of 18 of his essays (see SOURCES.md).

Search academic papers across multiple sources at once (OpenAlex, Semantic Scholar, arXiv/Valyu, Perplexity) from a single request, then cross-validate and cross-align the results into one cited synthesis report. Use this whenever the user wants to find papers / literature / prior work / studies on a topic and would benefit from MORE THAN ONE source — i.e. they want it comprehensive, corroborated, triangulated, or cross-checked, or they say things like "多源搜论文", "交叉搜论文", "跨源检索", "find papers o...

Find academic papers and read them — searches real scholarly databases for prior work on a topic, then resolves and deep-reads the open-access PDF of any paper you pick. Use this skill WHENEVER the user wants to discover literature or read a study: "find recent papers on X", "what does the research say about Y", "find sources for my thesis on Z", "search for prior work on ...", or "summarise / read / extract the findings from this paper / this arXiv id / this DOI / this PDF". Trigger even whe...

Consult the I Ching — cast hexagrams, interpret the Changes, engage the living oracle where bronze-age bones meet silicon probability matrices.

Apply the ScholarEval framework to systematically evaluate scholarly and research work using peer-reviewed assessment criteria across multiple quality dimensions.

Deep, multi-layered analysis of scientific papers and technical documents. Generates stratified readings from plain-language summary through graduate-level technical breakdown to critical synthesis. Use when asked to analyze, explain, critique, or engage deeply with scientific papers, preprints, patents, or technical research. Triggers on: 'analyze this paper', 'explain this research', 'break down this study', 'what does this paper say', 'help me understand this research', or any request for ...

Conducts multi-turn iterative deep research on specific topics within a codebase with zero tolerance for shallow analysis. Use when the user wants an in-depth investigation, needs to understand how...

Answers questions about a code repository using source file analysis. Use when the user asks a question about how something works, wants to understand a component, or needs help navigating the code...

Generate comprehensive market opportunity analysis with TAM/SAM/SOM

>

>

Expert web researcher using advanced search techniques and

An uncompromising Academic Research Engineer. Operates with absolute scientific rigor, objective criticism, and zero flair. Focuses on theoretical correctness, formal verification, and optimal impl...

When the user wants to optimize post-signup onboarding, user activation, first-run experience, or time-to-value. Also use when the user mentions \"onboarding flow,\" \"activation rate,\" \"u...

Implements a systematic framework for discovering, documenting, and validating

"Provides Finding and maintaining competitive advantage in trading systems"

论文全流程集大成助手（Academic paper all-in-one assistant and conductor）。覆盖学术论文从选题到投稿 的 8 个阶段（选题质询 → 文献调研 → 写作 → 制图 → 润色去AI味 → 审稿 → 答辩 → 投稿）。它自己就能动手： 写作（引言/方法/讨论/全文草稿）、润色与表达改写、基础审稿与自查、规划全程路线图，都直接做。 当某个环节有更强的专门 skill 时（深度系统性文献调研、框架图、中文破折号 hard gate 级去AI味、 学校 DOCX 模板交付、本地 .bib 检索、多 reviewer 深度评审），它会推荐并把产物交接过去。既动手又动嘴。 中文触发：写论文、帮我写引言/方法/讨论、润色这段、读起来像 AI、论文全流程、从头到尾写论文、 下一步该做什么、论文路线图、论文总调度、把这些阶段串起来。 English triggers: write my paper, draft a section, polish this, sounds too much like AI, end-to-end paper help, ...

Decomposition playbook + anti-temptation rules for an orchestrator profile routing work through Kanban. The "don't do the work yourself" rule and the basic lifecycle are auto-injected into every kanban worker's system prompt; this skill is the deeper playbook when you're specifically playing the orchestrator role.