发现 Skills

Detect and contain WordPress compromises, then harden the install against re-entry. Covers webshell detection across the Sid Gifari, WSO, FilesMan, b374k and c99 families, backdoored mu-plugins, malicious admin accounts, and shared-hosting lateral-movement defense. Invoke when a WordPress site shows unexpected files, suspicious admin accounts, defaced pages, or when hardening a fresh install on shared hosting.

Use when executing implementation plans with independent tasks in the current session

Download large datasets from NCBI efficiently using history server, batching, and rate limiting. Use when performing bulk sequence downloads, handling large query results, or production-scale data retrieval.

Integrate multiple scRNA-seq samples/batches using Harmony, scVI, Seurat anchors, and fastMNN. Remove technical variation while preserving biological differences. Use when integrating multiple scRNA-seq batches or datasets.

DNA methylation analysis with methylKit in R. Import Bismark coverage files, filter by coverage, normalize samples, and perform statistical comparisons. Use when analyzing single-base methylation patterns, comparing samples, or preparing data for DMR detection.

Analyze single-cell TCR and BCR data integrated with gene expression using scirpy. Use when working with 10x Genomics VDJ data alongside scRNA-seq or when integrating immune receptor information with cell state analysis.

Multi-round character casting orchestrator for the Horus movie pipeline. Extracts characters from scripts, discovers reference actors, generates identity images, and assigns voices. Runs as Phase 2.5 in create-movie.

Discover new movies using TMDB API with Federated Taxonomy integration. Recommends movies based on genres, directors, and HMT bridge attributes. Can be invoked via /dogpile movies.

Architecture skill with three modes: review existing codebase architecture (A1-A9 dimensions), create Architecture Decision Records, or design new systems from requirements. Modes: --mode review [path], --mode adr, --mode design. Uses CodeSift for module discovery, dependency mapping, structural metrics, and temporal coupling detection.

Checks consistency between memory files and the code. Detects orphans (memory files missing from MEMORY.md), ghosts (references in MEMORY.md pointing to non-existent files), and drifted values (constants in the code vs values in the docs). Use after a long session, before an audit, or when you suspect an inconsistency.

Use when implementing any feature or bugfix, before writing implementation code

Create comprehensive ontological documentation for Grey Haven systems — extract domain concepts from TanStack Start and FastAPI codebases, model semantic relationships, generate visual representations of system architecture, and document business domains. Use when onboarding, documenting architecture, or analyzing legacy systems.

Comprehensive Claude Code plugin auditing skill for validating structure, detecting deprecated patterns, and recommending best practices based on the latest changelog. Use when auditing plugins, checking for deprecations, validating plugin structure, preparing plugins for release, or ensuring compatibility with recent Claude Code versions. Triggers: 'audit plugin', 'check plugin health', 'validate skill', 'plugin deprecation', 'changelog compatibility', 'plugin best practices'.

Unified Memory Contract for Flowbaby integration. Defines when and how to retrieve and store memory. Load at session start - memory is core to agent reasoning, not optional.

Guide for showing user feedback using app_feedback package - toasts, snackbars, dialogs, bottom sheets (project)

Scoring system for opportunity hygiene, win likelihood, and inspection prioritization.

Framework for packaging competitive updates into executive-ready narratives and action plans.

Template for documenting co-built solutions, integrations, and GTM motions with partners.

Use when defining nurture pacing, triggers, and suppression logic.

Use when crafting or refreshing pages to align copy, UX, and conversion goals with target keywords.

Use to summarize persona goals, pains, proof points, and questions for call prep.

Use when structuring positioning statements, message houses, and proof libraries for GTM teams.

Creating algorithmic art using p5.js with seeded randomness and interactive parameter exploration. Use this when users request creating art using code, generative art, algorithmic art, flow fields, or particle systems. Create original algorithmic art rather than copying existing artists' work to avoid copyright violations.

[Deprecated: use baoyu-image-cards] Generates Xiaohongshu (Little Red Book) image card series with 12 visual styles, 8 layouts, and 3 color palettes. Breaks content into 1-10 cartoon-style image cards optimized for XHS engagement. Use when user mentions "小红书图片", "XHS images", "RedNote infographics", "小红书种草", "小绿书", "微信图文", "微信贴图", or wants social media infographic series for Chinese platforms.