Research & Search
Research agent skills teach AI agents to gather and synthesize information properly: literature reviews, competitive analysis, web research with citations, and structured summaries. Each SKILL.md encodes a method, not just a prompt, so results stay consistent across runs.
- ferroxlabs BundleTrue Crime Writer
Complete true crime writing guide covering research methodology, FOIA requests, interview techniques, ethical considerations, victim sensitivity, narrative structure, fact-checking protocols, and legal liability awareness. Use when the user asks about true crime writer, related techniques, best practices, or needs guidance in this domain. Do NOT use when the request is outside the scope of true crime writer or requires a different specialized skill.
Audited 618 - newmindsgroup SkillWiki Changelog
Generate structured changelogs from git history. Use when user asks \"what changed recently\", \"generate a changelog\", \"summarize commits\" or user wants to understand recent development activity.
Audited 2 - ferroxlabs BundleNeighborhood Researcher
Comprehensive methodology for researching and evaluating neighborhoods before purchasing property, covering school district evaluation, crime statistics analysis, walkability and transit scores, future development plans, property value trends, community amenities assessment, noise evaluation, commute analysis, and demographic research tools. Use when the user asks about neighborhood researcher or needs help with related topics. Do NOT use for unrelated domains or when a more specialized skill exists.
Audited 618 - adu2021 SkillChaining Evidence Rl
Train search agents using citation-aware rubric rewards that decompose complex questions into verifiable single-hop facts. Agents learn to chain evidence through explicit source citations, preventing hallucinations and shortcut exploitation. Citation-aware Group Relative Policy Optimization (C-GRPO) combines rubric and outcome rewards, enabling agents to solve multi-hop reasoning tasks with high factual grounding and transparency.
Audited 7 - nota-america SkillConstitutional AI
Anthropic's method for training harmless AI through self-improvement. Two-phase approach - supervised learning with self-critique/revision, then RLAIF (RL from AI Feedback). Use for safety alignment, reducing harmful outputs without human labels. Powers Claude's safety system.
Audited 93 - nota-america BundlePytorch Lightning
High-level PyTorch framework with Trainer class, automatic distributed training (DDP/FSDP/DeepSpeed), callbacks system, and minimal boilerplate. Scales from laptop to supercomputer with same code. Use when you want clean training loops with built-in best practices.
Audited 93 - undermybelt SkillAI Engineering From Scratch
AI engineering curriculum reference.
Audited 11 - ferroxlabs BundleAb Test Design Data Analysis
Designs an A/B test from scratch. Defines the hypothesis, calculates required sample size, specifies randomization method, sets success metrics and minimum detectable effect, and produces a results interpretation template. Use when the user wants to set up a controlled experiment to test whether a change produces a measurable improvement. Do NOT use for interpreting completed test results (use hypothesis-testing), analyzing existing data without experimentation (use correlation-analysis), or survey design (use survey-design in the research cluster).
Audited 618 - undermybelt SkillEcc Scholar Evaluation
Structured scholarly-work evaluation for papers, proposals, literature reviews, methods sections, evidence quality, citation support, and research-writing feedback.
Audited 11 - nota-america BundleLlama Factory
Expert guidance for fine-tuning LLMs with LLaMA-Factory - WebUI no-code, 100+ models, 2/3/4/5/6/8-bit QLoRA, multimodal support
Audited 93 - cslawyer1985 BundleLegal Citation
根据《法学引注手册(第二版)》生成符合规范的中文/外文学术脚注(引注)。接收文字、文献片段、图片、网页链接等任意输入,自动提取作者/标题/出处等要素,判断文献类型(中文图书、期刊论文、报纸、古籍、网络文献、未发表/会议/学位论文、法律法规、规范性文件、官方文件、司法案例、统计数据、英文文献、法德意俄日文献等),路由到对应规则并生成脚注。每次输出含主脚注、引领词、可选写法、再次引用简写形式,并在信息缺失时明确提示。This skill should be used whenever the user provides any document, citation fragment, screenshot, or URL and asks to generate, format, or check a legal-academic footnote/citation (脚注/引注/引用格式).
Audited 926 - cslawyer1985 BundleHilegal Compliance
This skill should be used when users ask about cross-border compliance, overseas legal compliance, foreign law research, corporate credit checks, global pricing queries, compliance consultation, or any compliance-related questions for international business. It supports queries in Chinese and English, covering compliance consulting, compliance Q&A, extraterritorial law research, corporate credit queries, and global price lookups. 当用户询问跨境合规、海外合规、域外法查明、企业征信查询、全球价格查询、合规咨询、 合规问答等需求时,应使用此技能。支持中英文双语,覆盖跨境企业海外运营的各类合规场景。
Audited 926 - findscripter SkillCitation Management
当撰写论文/学位论文需要系统管理参考文献时使用;做学术检索、元数据提取、引文校验并产出规范 BibTeX 与校验报告;不适用于正文叙事写作或非学术文档排版。触发词:BibTeX、DOI、引文管理、参考文献、PubMed
Audited 11 - cslawyer1985 BundleLegal Citation Verify
法规引用验证技能。当用户写文章、报告时引用了法律法规、政策文件、司法解释等,需要对引用的准确性、有效性、合理性进行验证时触发。覆盖六大验证组(来源溯源、元数据核验、效力与版本、法理一致性、适用条件、引用质量)共18项检查,支持法律、行政法规、部门规章、司法解释、政策文件、地方性法规六类引用的差异化验证管道。核心工具链为北大法宝(pkulaw)、华宇元典(yuandian)、互联网检索(WebSearch),采用引用归并+三阶段渐进验证架构。
Audited 926 - haibarakiku BundleAcademic Translator
Expert academic translator with 15+ years experience in scientific paper translation, language editing, and publication preparation
Audited 6 - haibarakiku BundleAnimal Experimenter
Senior animal experiment technician with 10+ years experience in rodent and small animal research. Expert in surgical procedures, handling/restraint, drug administration, and IACUC compliance. Specializes in
Audited 6 - onfire7777 BundleScientific Brainstorming
Research ideation partner. Generate hypotheses, explore interdisciplinary connections, challenge assumptions, develop methodologies, identify research gaps, for creative scientific problem-solving.
Audited 17 - onfire7777 BundleCommunication Storytelling
Use when transforming analysis/data into persuasive narratives—presenting to executives, explaining technical concepts to non-technical audiences, creating customer-facing communications, writing investor updates, announcing changes, turning research into insights, or when user mentions "write this for", "explain to", "present findings", "make this compelling", "audience is". Invoke when information needs to become a story tailored to specific stakeholders.
Audited 17 - yogsoth-ai SkillExplore Resume
Understand the user's background comprehensively — technical stack, project experience, research experience, publications, research directions. Allows user to express interest beyond their resume. Execute once only, never re-run.
Audited 512 - yogsoth-ai SkillCreative Ideation
Creative Generation Engine — transforms research hypotheses into diverse solution spaces via 10 parallel creativity campaigns spanning structural, analogical, destructive, and combinatorial methods.
Audited 512 - yogsoth-ai BundleScope Calibration
Strategy: Adjust research question scope — zoom in/out until the scope is appropriate
Audited 512 - yogsoth-ai SkillSystematic Survey
Exhaustive PRISMA-style literature survey — comprehensive coverage of all related work on a specific question. Multi-stage screening, citation chaining, quality assessment, and structured data extraction. Use when the user needs to demonstrate complete literature coverage or conduct rigorous gap analysis.
Audited 512 - yogsoth-ai BundleFramework Matching
SOP: Match the most suitable RQ framework based on research type
Audited 512 - yogsoth-ai SkillScope Clarification
Structured questioning SOP to determine research boundaries, depth, and breadth. Used during spec generation.
Audited 512 - yogsoth-ai BundleTimeline Projection
Extrapolate research landscape timelines using trend analysis and milestone projection
Audited 512 - yogsoth-ai BundleDecomposition Formulation
Strategy: decompose a complex research question into a hierarchy of independently answerable sub-questions
Audited 512 - yogsoth-ai BundleLadder Quality Order
Loss-2 judge (codex role). Over one topic's 6 shuffled research-design samples, pairwise-rank by quality using the D1–D5 standard. Emit the pairwise log; the harness computes the order and the ladder verdicts. Judge quality difference, never against academic standards.
Audited 512 - charlieviettq BundleHum Source Criticism
"Evaluate source credibility using primary/secondary classification, internal/external criticism, triangulation, and misinformation detection. Use this skill when the user needs to assess whether information is trustworthy, evaluate research sources, fact-check claims, or detect misinformation — even if they say 'can I trust this source', 'is this real', 'how reliable is this data', or 'fact-check this for me'.".
Audited 27 - ashfordeou BundleIn Flight Engine Relight Test
Use when you must reduce an in-flight engine restart demonstration: fit the least-squares windmill N2 regression against true airspeed from the windmill survey, read the minimum relight airspeed where the fitted N2 line crosses the required relight threshold, summarize the starter-assisted relight time-to-idle samples against the type-data limit per altitude band, and combine the band verdicts with the airspeed into the overall restart-demonstration verdict. Produces the regression slope, intercept and R-squared, the minimum relight airspeed in m/s, per-band time-to-idle statistics and verdicts, and the combined PASS or FAIL verdict. Trigger: in-flight relight test, engine restart demonstration, windmill N2 survey, relight airspeed, time-to-idle limit.
Audited 5 - tuyv BundleClassifying Tax Transactions
Content skill for classifying business bank transactions into US federal Schedule C (Form 1040) line items for sole proprietors and single-member LLCs disregarded for federal tax. Tax year 2025. Federal only. Supplies the Tier 1 deterministic vendor pattern library, Tier 2 conservative-default table, and refusal catalog. MUST be loaded alongside the tax-workflow-base skill which provides the three-state contract, citation discipline, structured-question form, and reviewer-attention output spec. This skill alone supplies rules but no workflow; the base alone supplies workflow but no rules.
Audited 1 - revfactory BundleAcademic Writing Style
A specialized skill that systematically guides academic thesis style, structure, and citation rules. Used by writing-coach and proofreader agents to improve academic writing quality and verify format compliance. Automatically applied in contexts such as 'academic style', 'thesis structure', 'citation format', 'APA style', 'argumentation structure', 'scholarly expression'. However, running plagiarism detection software (Turnitin) and journal submission on behalf of the user are outside the scope of this skill.
Audited 1.3k - revfactory BundleTam Sam Som Calculator
Methodology for systematically calculating market size (TAM/SAM/SOM). Use this skill for 'TAM SAM SOM calculation', 'market size estimation', 'market sizing', 'market opportunity analysis', 'addressable market', and other market sizing tasks. Note: purchasing paid market research reports and directly collecting census data are outside the scope of this skill.
Audited 1.3k - revfactory BundleRegulation Knowledge Base
A structured knowledge base of key regulatory laws. The 'law-mapper' and 'gap-analyst' agents must use this skill's law database and mapping methodology when identifying applicable laws by industry and mapping obligations. Used for 'applicable law verification', 'regulatory mapping', 'obligation extraction', etc. Note: Full compliance orchestration or remediation planning is outside the scope of this skill.
Audited 1.3k - ffsshhttiikk SkillUX
User experience design and research
Audited 3 - npbuilds BundleCounterfactual Prober
Active disconfirmation. Given a synthesized brief, generate "if the conclusion were false, what should we observe in the world?" predictions and search for those signatures. Stronger than passive adversarial search ("X debunked"), which only finds what others have already written. Plugged into Spelunker Phase 5 between the existing adversarial search and backpropagation. Triggered in `deep` mode by default; optional in `standard`.
Audited - ulpi-io BundlePr Plan
Plan an open source PR contribution. Takes pr-research output and produces scope, acceptance criteria, and risk assessment. Triggers: "pr plan", "contribution plan", "plan PR", "plan contribution".
Audited 15
Frequently asked questions
What are Research & Search agent skills?
Research agent skills teach AI agents to gather and synthesize information properly: literature reviews, competitive analysis, web research with citations, and structured summaries. Each SKILL.md encodes a method, not just a prompt, so results stay consistent across runs.
Which Research & Search skills are most installed?
Popular Research & Search skills on SkillMD right now include counterfactual-prober, classifying-tax-transactions, ux. Rankings shift as installs change; sort this page by "Most installs" for the live list.
Do Research & Search skills work with Claude Code and Cursor?
Yes. Every skill here ships as a SKILL.md file, an open format that works in Claude Code, Claude.ai, Cursor, Codex, Windsurf, and 60+ other agents. Install one with npx get-skill add <owner>/<name>, or copy the file into your agent's skills directory.