hunt enforces diagnostic discipline by blocking fixes until root cause is identified and stated with supporting evidence, reducing patch-induced regressions.

comfyui-nodes-dev covers server-side V3 custom node development in Python, with schema-based I/O, extension registration, and UI preview helpers. API adapter choice between stable v0_0_2 and latest is required upfront.

logo-creator orchestrates iterative ai logo design via gemini image generation, background removal, and svg vectorization. discovers user requirements, generates 20 variations, collects feedback, and delivers cropped, transparent, and vector-format assets.

web-research delegates multi-subtopic investigations to parallel subagents, aggregates findings from local files, and synthesizes cited reports. execution relies on explicit planning and file-based communication between agents.

runcomfy-cli wraps hundreds of ai model endpoints (image generation, video, editing, face-swap, upscaling) behind a single command-line interface with pluggable auth, multiple polling modes, and json output for scripting.

clerk-react-patterns covers @clerk/react setup for SPAs with Vite/CRA, including ClerkProvider initialization, useAuth/useUser/useClerk hooks, protected routes via React Router, and token handling for API calls.

role-creator scaffolds custom codex agent toml files with required metadata, enforces input validation before write, and provides guardrails for path resolution and syntax checking.

google-agents-cli-observability configures multi-tier tracing and logging for deployed ADK agents across cloud trace, bigquery, and third-party platforms, with terraform-managed infrastructure provisioning and privacy-preserving defaults.

shadcn covers component selection, composition patterns, cli workflows, and chat ui primitives. enforces strict rules on spacing, validation, icons, and semantic colors. project-aware and registry-conscious but lacks actionable trigger phrases for real-world scenarios.

mcp-oauth-cloudflare implements dual oauth roles (client and server) for mcp servers on cloudflare workers with google sign-in. covers security practices including csrf protection, token lifecycle management, and documented error prevention.

react-router-data-mode covers building react applications with data loaders and actions via createBrowserRouter and RouterProvider. includes patterns for forms, mutations, optimistic ui, and pending states without framework plugins.

conventional-commit standardizes git history using the conventional commits spec, mapping 12 commit types to semantic versioning and enforcing consistent message formatting with scope, body, and breaking change footers.

ctf-ai-ml covers model weight attacks, adversarial example generation, llm jailbreaking, and model extraction via quick-start commands and cross-referenced technique docs. includes pivot rules for crypto and reverse engineering.

contacts manages a markdown-based vault of enriched person profiles, pulling from 7 sources via inngest or manual entry. supports vip tracking, roam archive integration, and ongoing monitoring of key collaborators.

wizard generates interactive bash scripts that guide humans through tedious manual procedures (third-party setup, migrations, state transitions) with progress tracking, confirmations, and automated credential storage.

debugging-wizard provides a five-phase systematic methodology for isolating and fixing bugs, covering tool usage across python, javascript, and go, with explicit constraints on safe debugging practices.

powershell-windows documents critical syntax patterns for windows powershell including operator parentheses, null checking, error handling, and file path construction. covers common mistakes with actionable fixes and a usable script template.

cmux-browser automates browser interactions within cmux webviews via snapshot-based element references and state-change waits. covers navigation, form filling, clicking, and extraction on wkwebview surfaces.

system-architect transforms requirements into justified technical designs by extracting functional and non-functional requirements, selecting architectural patterns, mapping nfr categories to decisions, and defining components with clear boundaries and interfaces.

expect covers browser verification of code changes using playwright tools. verifies ui behavior in real browsers before marking work complete, handling session reuse and parallel execution patterns.

fine-dust-location queries korean air quality data (pm10, pm2.5, integrated grade) by region or station name via k-skill-proxy report endpoint, with candidate resolution for ambiguous locations.

webapp-testing provides playwright-based automation for local web application testing, with helper scripts for server lifecycle management and a decision tree for static vs dynamic content handling.

customer-journey-mapping-workshop guides product managers through structured discovery by posing five adaptive questions about personas, scenarios, phases, emotions, and opportunities, generating a visual artifact with ranked improvement priorities tied to specific pain points.

shopify covers app, extension, and theme development across checkout, admin, and pos surfaces using graphql apis, cli tooling, and liquid templating. includes setup workflows, code examples, and deployment patterns.

playwright-visual-testing provides browser automation and visual regression testing via playwright mcp server, covering cross-browser validation, form testing, responsive design checks, and e2e workflows with extensive tool reference and workflow patterns.

mapbox-data-visualization-patterns provides reusable code patterns for choropleth maps, heat maps, 3d extrusions, and data-driven styling on mapbox. covers color accessibility and basic error handling for missing data properties.

video covers ai-driven video production workflows including programmatic frameworks (hyperframes, remotion), ai generation models (veo, sora, runway, kling), avatar synthesis (heygen, synthesia), and repurposing tools. addresses demo, explainer, social, and batch video scenarios with tool selection guidance.

aiconfig-create guides users through LaunchDarkly AI config setup, covering shell creation, variation configuration, and verification. emphasizes critical targeting step before SDK evaluation to avoid false-positive failure diagnostics.

add-feature guides developers through extending browser games by planning feature scope, reading existing architecture, implementing with strict patterns (eventbus, constants, state), and verifying integration without breaking existing systems.

code-reviewer automates pull request analysis across six languages, detecting security risks, code quality violations, and complexity issues. generates structured review reports with severity-based verdicts and prioritized action items.

langchain4j-ai-services-patterns teaches declarative ai service construction via java interfaces, annotations, memory management, and tool integration. covers type-safe patterns with structured outputs and multi-user scenarios.

airunway-aks-setup guides deployment of ai runway on existing aks clusters through six sequential phases: cluster verification, controller installation, gpu assessment, provider setup, model deployment, and validation. includes cost warnings and partial-resume capability.

git-troubleshooting covers recovery techniques for lost commits, merge conflicts, detached HEAD states, corrupted repositories, and common error scenarios. provides concrete commands and multi-step procedures for diagnosis and remediation.

arize-evaluator covers designing, creating, and running llm-as-judge evaluators on arize. focuses on troubleshooting via ax command failures and enforces credential hygiene through platform integrations rather than filesystem search.

scikit-learn-best-practices covers workflow fundamentals like data splitting, feature scaling, pipelines, cross-validation, and model evaluation with working code examples. lacks operational guidance on failure modes, deployment specifics, and when to deviate from defaults.

session-logs provides jq recipes to search, filter, and analyze conversation history stored in append-only jsonl files, enabling retrieval of prior context across multiple chat sessions.

data-cleaning-pipeline covers systematic handling of missing values, outliers, duplicates, and data standardization using pandas and sklearn. includes practical python examples, quality metrics, and validation checks but lacks explicit error recovery strategies.

feishu-cli-chat covers reading message history, searching conversations, managing group membership, and interacting with messages via reactions and pins. delegates new message sending and card construction to specialized skills.

javascript-testing-patterns covers unit, integration, and component testing across jest and vitest with concrete examples for mocking, async operations, and dependency injection. punts advanced scenarios to external references rather than completing the narrative.

fix diagnoses and repairs flaky playwright tests by running targeted reproduction steps, categorizing failures against a taxonomy, and applying targeted fixes with verification loops.

cli-developer addresses command-line tool construction with emphasis on ux validation, framework selection, and cross-platform testing. covers argument parsing, subcommands, help text, and startup performance.

angular-developer generates framework code with version-aware guidance, validates builds, and covers the full stack from components to testing. incomplete execution guidance weakens procedural clarity.

third-party covers optimization strategies for analytics, ads, embeds, and widgets that degrade page performance. teaches resource hints, lazy-loading, self-hosting, and framework-level techniques to reduce main thread impact while retaining functionality.

seo-dataforseo wraps 79+ dataforseo mcp tools covering serp analysis, keyword research, backlink profiles, on-page audits, and llm mention tracking. includes cost guardrails and bulk domain comparison. requires separate extension install.

code-simplifier identifies verbose rust patterns in rtk and applies idiomatic transformations (iterators, string joins, option chaining, early returns) while preserving mandatory constraints like lazy-static regex, context error chains, and test modules.

notebooklm enables querying google notebooklm notebooks via claude code with source-grounded answers. manages authentication, notebook libraries, and multi-turn follow-ups through python script wrappers that handle environment setup automatically.

databricks-core covers cli authentication, profile management, resource enumeration, and data exploration via aitools. handles common pitfalls around positional arguments and shell session isolation.