发现 Skills

Deep design and implementation analysis with devil's advocate evaluation

Service decomposition, API Gateway with YARP, inter-service communication (gRPC, REST, message bus), distributed caching, health checks, circuit breakers

Tactical Blueprint for SAP Fiori Elements development. Focuses on metadata-driven UI development via CDS annotations and SAP Fiori Tools.

ASP.NET Core Minimal APIs, Controllers, dependency injection, middleware, background services, options pattern, health checks

CPI iFlow design patterns (content-based routing, splitter, aggregator), Groovy script patterns, API Management policies, Event Mesh topic design, OData consumption

Tactical Blueprint for ML Training, Deployment, and Monitoring. Focuses on production-grade PyTorch loops, distributed training, and MLOps observability.

Token streaming from LLMs, event streaming with astream_events, WebSocket agent patterns, and real-time UI updates

Vitest setup and configuration, React Testing Library patterns, component testing (render, fireEvent, waitFor), mock patterns (MSW for API mocking), Playwright E2E testing, snapshot testing strategy, and coverage configuration.

Provides AWS Lambda integration patterns for Java with cold-start optimization. Use when deploying Java functions to AWS Lambda, choosing between Micronaut and raw Java approaches, optimizing cold starts below 1 second, configuring API Gateway or ALB integration, or implementing serverless Java applications. Triggers include "create lambda java", "deploy java lambda", "micronaut lambda aws", "java lambda cold start", "aws lambda java performance", "java serverless framework".

Provides comprehensive code review capability for NestJS applications, analyzing controllers, services, modules, guards, interceptors, pipes, dependency injection, and database integration patterns. Use when reviewing NestJS code changes, before merging pull requests, after implementing new features, or for architecture validation. Triggers on "review NestJS code", "NestJS code review", "check my NestJS controller/service".

Generate and query structural indexes for Drupal projects. Parses services.yml, routing.yml, hooks, plugins, and entity types to build an automated dependency graph. Use for blast radius analysis, service dependency mapping, and structural queries. Commands include `/structural-index` to generate, and structural query prefixes via discover (e.g., `discover service:entity_type.manager`, `discover deps:AUTH`).

Configure and implement WiFi connectivity in ESP32 projects using ESP-IDF

Configure and implement WiFi connectivity in ESP32 projects using ESP-IDF

Agent boundary checklist. The most dangerous thing in AI products is not that something can’t be done, but that it’s done too far. This Skill helps you systematically define what an AI should do, should not do, and when it must stop and wait for a human. Based on the "AI rebuild product needs" tool card.

Generate developer handoff specs from a design. Use when a design is ready for engineering and needs a spec sheet covering layout, design tokens, component props, interaction states, responsive breakpoints, edge cases, and animation details.

Survey a module, file, or PR for missing test coverage. Produce a specific, non-hallucinated list of test cases that should exist, with enough detail that another developer (or agent) can implement them. Optionally open GitHub issues for each gap. Use when user says "find missing tests", "what's not tested?", or wants a test backlog for a piece of code.

Reference skill for Zoom REST API. Use after choosing an API-based workflow when you need endpoint selection, resource-management patterns, OAuth requirements, rate-limit awareness, or API error debugging.

Research a repo, then actually *run* the "how to set up the env / run tests / lint / type-check" commands, and produce a clean pair of agent-context files for the project — `AGENTS.md` (the cross-tool convention every coding agent reads) plus a repo-root `CLAUDE.md` that `@AGENTS.md`-imports it (Claude Code only reads `CLAUDE.md`, not `AGENTS.md`). If those files already exist, *improve* them rather than overwrite — re-research, re-verify, and reconcile each existing claim. Verified commands get a `✅ verified <date>` marker so downstream agents and human reviewers know they were actually executed, not guessed. Use this when the user wants to "set up AGENTS.md / agent context for this project", "this repo has no CLAUDE.md, create one", "the AGENTS.md is out of date, fix it", "bootstrap project context", or "/spike-init". The difference from Claude Code's built-in `/init`: `/init` analyzes the codebase and writes down the commands it *infers*; `/spike-init` runs those commands first and only writes the ones that actually worked.

Manages datasets, tables, and jobs in BigQuery, and integrates with BigQuery ML and Gemini for advanced data analytics and AI-driven insights. Use when you need to interact with BigQuery, run SQL queries, manage BigQuery resources, or leverage BigQuery's built-in ML capabilities. Also use when performing data analysis, ingesting data into BigQuery, or developing AI applications on BigQuery.

Guides the usage of the Gemini API on Agent Platform with the Google Gen AI SDK. Use when the user asks about using Gemini in an enterprise environment or explicitly mentions Vertex AI, Google Cloud, or Agent Platform. Covers SDK usage (Python, JS/TS, Go, Java, C#), capabilities like Live API, tools, multimedia generation, caching, and batch prediction.

Reverse-engineer “image-format PPT documents / slide screenshots / any image page” back into an editable .pptx. Enforced four-layer pipeline: replicate the background image + chroma-key (green-screen) extraction of the overall frame (skeleton) + chroma-key extraction of element icons/decorations + GPT visual text extraction, then compose. The text becomes real editable text boxes, and frame images/icons become movable images; frames are split into sub-parts only when the user explicitly requests it. Use when the user wants to “convert PPT images/screenshots into editable PPTX, restore image-based PPT to editable PPT, extract icons, replicate background, extract text, or restore layout.” Input: images. Output: an editable .pptx.

Use when the task requires automating a real browser from the terminal (navigation, form filling, snapshots, screenshots, data extraction, UI-flow debugging) via `playwright-cli` or the bundled wrapper script.

Performs structure-based virtual screening using AutoDock Vina 1.2 for molecular docking. Prepares receptor PDBQT files, generates ligand conformers, defines binding site boxes, and ranks compounds by predicted binding affinity. Use when screening chemical libraries against a protein structure to find potential binders.

Interact with GitHub using the `gh` CLI. Use `gh issue`, `gh pr`, `gh run`, and `gh api` for issues, PRs, CI runs, and advanced queries.