Skill catalog

Optimizes or restructures CLAUDE.md/AGENTS.md with progressive disclosure and zero information loss. Use when the user asks to optimize, audit, 精简/瘦身/重构, or split instruction files, asks for their best practices (最佳实践),…

Schema.org structured data audit and generation optimized for AI discoverability — detect, validate, and generate JSON-LD markup

When the user wants open source strategy, OSS commercialization, or open source growth. Also use when the user mentions "open source strategy," "OSS strategy," "open source commercialization," "open source to paid," "op…

Convert PRDs to beads for ralph-tui execution. Creates an epic with child beads for each user story. Use when you have a PRD and want to use ralph-tui with beads as the task source. Triggers on: create beads, convert pr…

Query Windsor.ai business data across marketing, sales, CRM, ecommerce, finance, and analytics connectors. Use when users need dashboards, reports, data visualization, schema exploration, or connector-backed test data f…

Add a new gameplay feature to an existing browser game following its architecture patterns. Use when the user says "add a feature", "add double-jump", "add a power-up", "add a leaderboard", or describes a specific game…

[BETA] Start the dev server, open the feature in a browser, and iterate on improvements together.

Publication-figure overlay for draw.io. Use instead of drawio whenever the diagram is for a paper, thesis, dissertation, journal, conference, IEEE/ACM submission, manuscript, camera-ready, Word/LaTeX figure, or other pu…

When the user wants to plan, implement, or optimize student and education discount programs. Also use when the user mentions "student discount," "education discount," "student plan," "for students," ".edu discount," "ac…

Execute ready image briefs and edits, including product or persona images and batch variants. Use resolved references and return image assets with saved run results.

Train Mixture of Experts (MoE) models using DeepSpeed or HuggingFace. Use when training large-scale models with limited compute (5× cost reduction vs dense models), implementing sparse architectures like Mixtral 8x7B or…

Archives completed changes and merges specification deltas into living documentation. Use when changes are deployed, ready to archive, or specs need updating after implementation. Triggers include "openspec archive", "a…

Expert in Python development with best practices across web, data science, and automation

Use when one look must hold across Scenario generations: one character across scenes, a turnaround, or a video animated from its references, one product across angles, one style across icon sets, a character from an upl…

Master 20 SEO & GEO skills for Claude Code agents - keyword research, content optimization, technical audits, rank tracking with CORE-EEAT and CITE frameworks

Fast structured generation and serving for LLMs with RadixAttention prefix caching. Use for JSON/regex outputs, constrained decoding, agentic workflows with tool calls, or when you need 5× faster inference than vLLM wit…

No description provided.

Monitor a pull request through review and CI. Use when the user asks to monitor, watch, or babysit a PR.

Open-source embedding database for AI applications. Store embeddings and metadata, perform vector and full-text search, filter by metadata. Simple 4-function API. Scales from notebooks to production clusters. Use for se…

Enable multiple Claude Code instances to discover each other and exchange messages in real-time via a local broker daemon and MCP server.

Looks up implementation details in the latest Cloudinary docs via the relevant llms.txt file. Use when building code or answering questions relating to image or video uploads, asset management and search, the Media Libr…

Extract structured data from LLM responses with Pydantic validation, retry failed extractions automatically, parse complex JSON with type safety, and stream partial results with Instructor - battle-tested structured out…

Use when writing or reviewing Kotlin code that stores CoroutineScope, launches from init/non-suspending APIs, calls runBlocking, or catches broad exceptions around suspend calls.

Merge multiple fine-tuned models using mergekit to combine capabilities without retraining. Use when creating specialized models by blending domain-specific expertise (math + coding + chat), improving performance beyond…

GPU-accelerated data curation for LLM training. Supports text/image/video/audio. Features fuzzy deduplication (16× faster), quality filtering (30+ heuristics), semantic deduplication, PII redaction, NSFW detection. Scal…

Use when quickly generating a single OpenCLI command from a specific URL and goal description. 4-step process — open page, capture API, write TS adapter, test. For full site exploration, use opencli-explorer instead.

Setup Sentry in React apps. Use when asked to add Sentry to React, install @sentry/react, or configure error monitoring, error boundaries, session replay, or browser tracing for React applications.

Accelerate LLM inference using speculative decoding, Medusa multiple heads, and lookahead decoding techniques. Use when optimizing inference speed (1.5-3.6× speedup), reducing latency for real-time applications, or depl…

Creates, deploys, and troubleshoots BigQuery User-Defined Functions (UDFs) and Remote Functions across SQL, JavaScript, Python, and Cloud Run endpoints. Use when writing custom scalar or aggregate functions in BigQuery,…

Print a summary of the Cyrus setup and offer to start the agent.

Elasticsearch development best practices for indexing, querying, and search optimization

Web content extraction via Jina AI Reader API. Three modes: read (URL to markdown), search (web search + full content), ground (fact-checking). Extracts clean content without exposing server IP.

Expert guidance for fine-tuning LLMs with LLaMA-Factory - WebUI no-code, 100+ models, 2/3/4/5/6/8-bit QLoRA, multimodal support

Neo4j Python Driver v6 — driver lifecycle, execute_query, managed and explicit transactions, async (AsyncGraphDatabase), result handling, data type mapping, error handling, UNWIND batching, connection pool tuning, and c…

Monitorar manifests e arquivos estruturais para criar ou atualizar .specsfy/STACK.md de forma aditiva. Use quando o monitor de contexto apontar stack pendente; quando framework, linguagem, runtime, pacote estrutural, fe…

Provides guidance for mechanistic interpretability research using TransformerLens to inspect and manipulate transformer internals via HookPoints and activation caching. Use when reverse-engineering model algorithms, stu…

Guides the agent through migrating an existing Capacitor app project from CocoaPods to Swift Package Manager (SPM) for iOS dependency management. Covers prerequisite checks, inventorying installed Capacitor plugins, bac…

Autonomous multi-page site builder using a baton-passing loop. Each iteration reads a task from .design/next-prompt.md, generates a page in HTML/Tailwind, integrates it into the site, verifies visually, then writes the…

Provides PyTorch-native distributed LLM pretraining using torchtitan with 4D parallelism (FSDP2, TP, PP, CP). Use when pretraining Llama 3.1, DeepSeek V3, or custom models at scale from 8 to 512+ GPUs with Float8, torch…

Plan short-form post-edit decisions from A-roll, B-roll, scripts, and reference videos. Use this when the goal is not generic video analysis or rendering, but deciding how to cut a social video beat by beat, including w…

The Facebook strategy playbook for 2026. Use when someone wants to "grow on Facebook," "fix my Facebook reach," "a Facebook content/Page strategy," "Facebook Reels," or how the Facebook algorithm works now. Page feed br…

High-performance RLHF framework with Ray+vLLM acceleration. Use for PPO, GRPO, RLOO, DPO training of large models (7B-70B+). Built on Ray, vLLM, ZeRO-3. 2× faster than DeepSpeedChat with distributed architecture and GPU…

Open Plannotator on the latest rendered assistant message and use the returned annotations to revise that message or continue.

Fan out N parallel workers, drain them, and return one report. Use for /swarm, 'swarm this', or parallel coverage, races, gauntlets, and exploration.

Generate genuinely beautiful, on-brand UI instead of generic "AI slop" — use whenever the user asks to build, design, style, or improve a UI, landing page, dashboard, app screen, or component, whenever a PRD/spec needs…

Remove AI writing patterns from prose. Use for docs, READMEs, blog posts, any non-code text that should sound human. Triggers: "unslop", "humanize", "make this sound human", "clean up this writing", "edit for voice", "d…

Use for "automate me", "create/update/refresh my -mode skill", "turn/capture my preferences or working style into a skill", or wanting agents to follow how the user works. Drafts or revises a personal -mode skill via pl…

Use when a developer wants to build or validate Block Kit layouts for Slack messages, modals, or Home tabs: message layouts, modals/forms/dialogs, Home tab interfaces, interactive buttons or menus, or modifying existing…

Use when investigating Jetpack Compose recomposition performance, skippable/restartable composables, composables.txt or compiler reports, Layout Inspector recomposition counts, back-writing snapshot state across phases,…

No description provided.