Build and maintain isolated DEV-only feature labs (demo/inspection routes that exercise one system in isolation).
Guide de recherche UX : interviews, tests utilisateurs, personas et analyse. Se déclenche avec "UX research", "recherche utilisateur", "interview utilisateur", "test util — from…
Structured development workflows using /brainstorm, /write-plan, and /execute-plan patterns. Transform ad-hoc conversations into systematic project execution with…
Load development-mode context for codebase work — engineering principles and environment conventions.
AI Agent / LLM 应用工程化方法论。指导如何把一个 AI 智能体应用"拆成哪几块 + 如何避免重走老路"：分层管道、规则优先+LLM兜底、多模型触点路由、Prompt 统一治理、超时/缓存/降级、评测闭环、多厂商 SDK 适配。Triggers on "AI agent 架构", "LLM 应用怎么拆", "规则还是大模型", "prompt…
Review completed dev sessions to decide whether work should be distilled into devlab-* assets (skills, CLI tools, MCP servers, loops, subagents, or knowledge cards).
前后端契约规范技能（contract 技能簇首个，定位介于 devlab-srv-* 与 devlab-web-* 之间）。约束大型前后端分离项目的接口/序列化契约、字段类型一致性、错误码与配置分层，提供契约校验与联调防错清单。Triggers on "前后端契约", "接口对不上", "序列化", "字段类型不一致", "contract", "api…
DAO 层通用 SQL 方言兼容性检查 + 系统性修复工作流。支持 MyBatis/JPA/MyBatis-Plus/SQLAlchemy 四种持久层框架，扫描源码识别 Oracle/PostgreSQL/MySQL 方言混用风险，自动修复 80% 常见陷阱，剩余 20% 由 Agent 辅助人工确认。七阶段流程 + adapter…
数据库数据层安全迁移专家技能。覆盖 schema 演进（加列/改约束/建索引）与数据搬迁的完整安全流程：备份→幂等检测→迁移→验证→失败回滚。当项目需要变更已有表结构、生产报 no such column / 唯一约束冲突、或需编写独立迁移脚本时使用。与 devlab-dao-sql-compat 互补：本技能管数据层面，后者管 SQL…
eval 自评测体系驱动的 AI Agent 生产体系方法论。以评测集为核心资产，建立"评测集组织 + Mock 隔离 + 标准化比对 + 自动评测脚本 + 回归门禁"闭环；L3 数据后端接 Langfuse（dataset/scores/LLM-as-judge），配套 evalctl CLI（已发布 0.2.x）。Triggers on "评测集",…
通用 spec 驱动的 AI 协作开发方法论。把需求转成 requirements/design/tasks 三件套提案作为事实源，经审查→细化→分任务执行→回归验证，全程 human-on-the-loop。适用于任何支持 spec 工作流的 AI IDE/Agent（Kiro spec 仅为一个 example）。Triggers on "spec…
后端 API 测试专家技能。支持 Java（JUnit）/ Python（pytest）/ Node.js（supertest + jest）后端项目的接口测试、Mock 服务、测试数据管理。
技术债全生命周期运维：审计分析\u2192建议报告\u2192确认后重构（微观+宏观）\u2192回归验证。覆盖从\u201c发现债务\u201d到\u201c安全消除\u201d的闭环。Triggers on \"\u91cd\u6784\", \"\u6280\u672f\u503a\",…
前端项目上下文理解规则资产（-context 家族首个成员）：双模式（模式 A 规则包被 devlab-context-bootstrap 文件契约委托 / 模式 B 独立四步渐进式理解），产出前端 context 画像（7 固定键 + 置信度标注），供 bootstrap 消费填五件套台账。
存量 Web 系统深度功能验收方法论（L1-L4）。registry 功能点驱动 + 唯一入口执行器 + 假成功嗅探 + 三对齐审计，覆盖盘点建模、用例固化、失败五分类路由、多会话认领协作。与 devlab-web-test-e2e（绿field 用例生成）分工：本技能面向已有系统的全功能点深度验收与回归资产管理。实战陷阱全量清单见…
ECharts 在 Astro/前端项目中的完整集成指南，包含 tree-shaking、IntersectionObserver 按需加载、自定义系列集成、拓扑图实现、雷达图配置、Astro Islands 数据传递、调试经验。
Web / 前端工程化能力层（Frontend Engineering Capability）。语言无关覆盖六域：Project Bootstrap、Build、Dependency、Packaging、Runtime/Environment、Engineering Convention。路由键 = 语言 × 框架 双维：TS 主线 / JS 兼容 /…
Use when building a browser/web extension from scratch, especially one that extracts, downloads, or enhances data from an existing web app.
前端 GlitchTip 错误追踪接入技能。凡是用户提到 GlitchTip、Sentry SDK、前端错误聚合、Source Map 上传、JS 错误追踪、Issue 管理时触发。框架无关，支持 Vue/React/Angular/Svelte/原生 JS。
前端 SkyWalking RUM 接入技能。凡是用户提到前端监控、RUM、skywalking-client-js、Web Vitals 采集、浏览器端链路追踪、前端错误上报时触发。框架无关，支持 Vue/React/Angular/Svelte/原生 JS。
Web E2E 测试专家技能。专注 Vue/React 前端项目的浏览器端到端测试能力，基于 Playwright + AI Agent 实现测试计划生成 → 代码生成 → 失败修复的完整闭环。
Web 视觉化操作（Web Visual Operations）：对 Web 应用做 Observe/Navigate/Interact/Inspect/Validate/Capture 的 day-2 操作手册。Router 层统一编排 Playwright（确定性）、Stagehand（AI…
Pipelines GitLab CI/CD — stages, jobs, runners, artifacts, environments et Auto DevOps. Se déclenche avec "GitLab CI", "gitlab-ci.yml", "pipeline GitLab", "runner GitLab" — from…
Hypothesis-first bug diagnosis. Verify the user's diagnosis before proposing alternatives. Keywords: bug, diagnose, debug, fix, hypothesis, root cause, crash, error, timeout,…
Find the root cause of a bug on a cliban ticket and prove it, logging the hypothesis ledger to the ticket as you go.
Convert informal English words, phrases, collocations, and sentence fragments into precise TOEFL-appropriate alternatives while preserving meaning and register.
Simulated academic peer review for economics papers, named in honour of Claude Diebolt. Creates independent referee agents who review your paper in isolation.
Проверка и подготовка диплома/ВКР к прохождению антиплагиата и ИИ-детектора (ИТМО и аналоги). Сканирует текст на признаки GPT-генерации по 10-пунктному чек-листу, выдаёт…
Guides writing a diploma thesis at PEF ČZU Praha. Use when the user asks to write, structure, cite, or review a diploma thesis, závěrečná práce, or any section thereof (úvod, cíl,…
Assistente academico para a disciplina DIR410346 (PPGD/UFSC, Prof. Diego Nunes, 2026.1, segundas-feiras).
Generate complex multi-page PDFs by running reportlab Python code directly via run_shell when shell_agent fails on document creation
Create a writing scaffold teaching the genre conventions specific to an academic discipline. Use when students write lab reports, historical essays, or other discipline-specific…
Turning research artifacts into actionable PM insight. Customer interviews, user research notes, support ticket reviews, sales call transcripts, survey data, in-app feedback, all…
Structures and writes discussion sections for academic papers and research reports. Use when writing a discussion section, interpreting research results, connecting findings to…
Write a letter to dispute an incorrect charge, bill, or record. Use when asked to dispute a credit-card charge, contest a bill or invoice, challenge a credit-report error, or…
Use when: Writing dissertations, theses, research projects, or individual chapters (Chapter 1-5). Follow strict academic structure for Cameroon-based research with quantitative…
Re-anchor research and no-assumptions discipline mid-session, then self-audit and correct the current work.
Research-driven critique and enhancement of Amelia architecture or ideas docs. Use with GPT-5.2 when given a design/architecture doc to analyze concepts, find sources, an — from…
Recherche académique dans les bases de données de doctrine juridique française, européenne et internationale.
Common conventions for writing Markdown documents that cite external sources (survey reports, paper explainers, and any deliverable that surfaces facts from outside).
MCP-powered domain research for requirements elicitation. Uses perplexity, context7, firecrawl, and other MCP servers to research domain knowledge, best practices, and industry…
Use when formulating clinical research questions (PICOT framework), evaluating health evidence quality (study design hierarchy, bias assessment, GRADE), prioritizing…
Drafts court filings — motions, memoranda of law, appellate briefs — as court-ready .docx, with Midpage research behind every citation.
Use STORM to research a topic, build a multi-perspective outline, and generate a cited long-form draft for human review.
Use when drafting an internal legal memorandum (research memo) analyzing a legal question for a supervising attorney, client, or internal team.
Maximizes knowledge discovery through Experimental and Formal Methods using short BIBIFI cycles. Use for Strong Inference, アノマリーからアブダクション, 競合仮説・予測表・排除表, 実験計画, 6分イテレーション, マイクロチケット,…
Build complex AI systems with declarative programming, optimize prompts automatically, create modular RAG systems and agents with DSPy - Stanford NLP's framework for syst — from…
Chambres Africaines Extraordinaires (CAE) / Extraordinary African Chambers in the Courts of Senegal (EAC) research, drafting, and analysis — the Hissène Habré case.
Generate complete visual ebooks where EVERY page is a full-page AI-generated image (cookbooks, manga, picture books, visual guides).
Research-before-coding workflow. Search for existing tools, libraries, and patterns before writing custom code.
Extraordinary Chambers in the Courts of Cambodia (ECCC, Khmer Rouge Tribunal) research, drafting, and analysis.
Verification-first methodology for the ECCC (Khmer Rouge Tribunal). Every citation is verified against eccc.gov.kh, the ECCC archive or legal-tools.org.
Use when designing or auditing the experimental program of an ECCV paper — benchmark selection that survives a September conference, matched-substrate baseline fairness in the…
A 股 ECM 律师专用案例检索 Skill。针对特定法律问题、交易结构、合规争点或尽调发现， 对多来源案例库进行检索、筛选、要点提炼和适用性分析。覆盖五类案例：中国裁判文书（最高法/高院/ 中院/基层法院判决、裁定、调解）、证监会行政处罚决定书、交易所纪律处分决定书、并购重组审核委员会…
A 股 ECM 律师专用法规检索 Skill。针对特定业务主题、尽调发现或方案设计需求， 对五级法规体系进行检索、时效性核验和条款摘录。覆盖五个效力层级：法律（全国人大及其常委会）、 行政法规（国务院）、部门规章（证监会、财政部、人民银行、国资委、商务部等部委令）、 规范性文件（证监会公告、各部委答复等）、交易所业务规则（上交所/深交所/北交所/新三板的…
A 股 ECM 律师专用法规深度研究 Skill。在已有法规清单基础上，对特定法律问题的法规体系 做深度研究：效力层级梳理、一般法与特别法关系、新法与旧法衔接、上下位法冲突识别与解决、 法规适用性分析。与 ecm-research-reg-search 的关系：reg-search 负责"找法规"，reg-study…
Review economics papers, working papers, theses, and submissions with referee-grade rigor and exhaustive verified comments.
Use when the identification argument is the bottleneck for an Economic Policy (EP) manuscript — causal identification of a policy effect, or parameter identification in a…
Education Data Architecture (EDA) data model deep dive: Contact-centric, Affiliations, Relationships, Account record types (Academic Program, Educational Institution, Sports…
Use when auditing an EDBT research-track submission for Microsoft CMT readiness, covering the cycle choice, the paper shape (Regular / Experiments-&-Analysis / Vision), the…