发现 Skills

Debug Django web applications with systematic diagnostic approaches. This skill covers troubleshooting Django-specific errors including TemplateDoesNotExist, ImproperlyConfigured, IntegrityError, migration conflicts, CSRF failures, N+1 query problems, and circular imports. Includes Django Debug Toolbar setup, ORM query logging, pdb/ipdb usage, shell_plus debugging, and comprehensive logging configuration. Provides four-phase methodology for root cause analysis and regression prevention.

Debug Docker containers, images, and infrastructure with systematic diagnostic techniques. This skill provides comprehensive guidance for troubleshooting container exit codes, OOM kills, image build failures, networking issues, volume mount problems, and permission errors. Covers four-phase debugging methodology from quick assessment to deep analysis, essential Docker commands, debug container techniques for minimal images, and platform-specific troubleshooting for Windows, Mac, and Linux.

Debug FastAPI applications systematically with this comprehensive troubleshooting skill. Covers async/await issues, Pydantic validation errors (422 responses), dependency injection failures, CORS configuration problems, database session management, and circular import resolution. Provides structured four-phase debugging methodology with FastAPI-specific tools including uvicorn logging, OpenAPI docs, and middleware debugging patterns.

Debug Rails issues systematically. Use when encountering ActiveRecord errors like RecordNotFound, routing issues, N+1 query problems detected by Bullet, asset pipeline issues, migration failures, gem conflicts, ActionController errors, CSRF token problems, or any Ruby on Rails application errors requiring diagnosis.

Systematically debug SwiftUI application issues. This skill helps diagnose and resolve SwiftUI-specific problems including view update failures, state-management bugs with @State/@Binding/@ObservedObject, NavigationStack/navigationLink issues, memory leaks caused by retain cycles, preview crashes, Combine publisher problems, and animation glitches. Provides practical Xcode debugger techniques, Instruments profiling tips, and LLDB commands tailored for iOS/macOS development.

Debug Vue.js 3 application issues systematically. This skill helps diagnose and resolve Vue-specific problems including reactivity failures with ref/reactive, component update issues, Pinia store state management problems, computed property caching bugs, Teleport/Suspense rendering issues, and SSR hydration mismatches. Provides Vue DevTools usage, console debugging techniques, Vite dev server troubleshooting, and vue-tsc type checking guidance.

Refactor Kubernetes configurations to improve security, reliability, and maintainability. This skill applies defense-in-depth security principles, proper resource constraints, and GitOps patterns using Kustomize or Helm. It addresses containers running as root, missing health probes, hardcoded configs, and duplicate YAML across environments. Apply when you notice security vulnerabilities, missing Pod Disruption Budgets, or :latest image tags in production.

Maintain and enforce constraint specifications for Android mobile applications. Triggered when users request adding, modifying, refactoring, auditing, or aligning the `android` rule set; applicable to rule-maintenance scenarios for Kotlin/Java native Android apps such as architecture layering, dependency injection, Compose/XML conventions, and secure storage.

Maintain and enforce .NET desktop application constraint specifications. Triggered when users request additions, modifications, refactoring, auditing, or alignment of the `dotnet-desktop` ruleset. Applicable to rules-maintenance scenarios for WPF, MAUI, and WinForms desktop applications, covering MVVM/MVP architecture, UI thread, local storage, auto-update, and related constraints.

Electron desktop application coding guideline assistant. Triggered when AI writes Electron main-process or preload code; it automatically loads the appropriate subset of rule files for the coding scenario to constrain code output. The renderer (frontend) follows the rules/frontend specification; this skill only constrains Electron main-process and preload-specific parts.

Automatically initializes an Electron desktop application project structure based on the renderer front-end framework. When starting a new project, after you specify the front-end framework (react/vue), it reads the corresponding rules and generates the main process, preload, and renderer process directory structures, configuration files, CI scripts, and produces an initialization report.

Maintain and enforce front-end/back-end collaboration constraints. Triggered when users request defining or adjusting API contracts, error-code mappings, integration workflows, release ordering, and rollback strategies. Especially applicable to scenarios that require joint cross-end consistency audits with `rules/frontend`, `rules/go-server`, `rules/dotnet-server`, and `rules/frontend-backend-collaboration.md`, and that must produce mergeable conclusions (including P0/P1 risk levels).

Frontend coding guideline assistant. Triggered when AI writes frontend code (pages, components, routes, state management, etc.). It automatically loads the corresponding subset of rule files based on application type (admin-console / wechat-h5 / miniprogram) and framework (Vue3 / React) to constrain code output. It can also be used to guide human developers to follow the standards.

A/B testing patterns for lead-generation sites. Client-side experiments, Google Optimize alternative, analytics integration. Use for conversion optimization.

Chief problem solver — investigations, audits, ad hoc multi-agent tasks, routing. Activates on '/cowboy', 'audit this', 'investigate', 'figure out what's wrong', 'check for conflicts', 'what happened here', 'handle this', or when work requires heterogeneous agents doing different things.

Produces a structured skill audit report after real use. Activates on '/skillaudit [skillname]', 'audit this skill', 'how did that skill do', 'let's get feedback on this skill'. Asks 6 standard questions, outputs a clean SKILLAUDIT_[name]_[date].md for the skills manager. Does NOT apply changes — that's the skills manager's job.

Loaded when user asks about TypeScript patterns or type safety

Converts Mini Apps from MiniKit (OnchainKit) to native Farcaster SDK. Use when migrating from @coinbase/onchainkit/minikit, converting MiniKit hooks, removing MiniKitProvider, or when a user mentions MiniKit, OnchainKit, or Farcaster SDK migration.

Build or refresh a high-dimensional user profile from `syft following` output. Use when Codex needs to turn a Syft CLI following list into reusable profile artifacts such as `following_topics.md`, raw interest or aversion blocks, and an interpretive `profile_summary.md`, especially when only Syft CLI output is available and richer dashboard or trace artifacts are missing.

Physical location awareness from FindMy. Use when user says "physical", "where am I", "location", "where is nat", or needs to check current physical location.

X-ray any Polymarket wallet — skill level, entry quality, bot detection, and edge analysis. Queries Polymarket's public APIs, no authentication needed. Inspired by @thejayden's "Autopsy of a Polymarket Whale" analysis.

Comprehensive performance audit workflow for web applications.

Orchestrate changing requirements without task sprawl by reconciling new asks against the active epic/spec, updating the source-of-truth brief before task edits, applying supersession precedence for newer approved requirements, deciding when to edit vs split vs create net-new tasks, and handling ambiguous founder requests with a decision-first decomposition flow.

Delegate multi-step execution to sub-agents with strict briefs, model routing, timeouts, and concise done-message contracts.