177 skills · sorted by relevance
Use Phoenix to record execution traces, detailed records of an AI application's operations, and investigate problems. Evaluate answers on datasets, compare prompts and models, and set up monitoring through a self-hosted server.
Use LangSmith to record LLM application inputs, outputs, and timing, investigate failures, evaluate answers against test datasets, and monitor production usage. It also guides automated quality checks and analysis of token usage and costs.
Get guidance on adapting LLMs through additional training with LLaMA-Factory. The skill covers a visual interface requiring no code, LoRA and QLoRA training methods, and documentation-based help with implementation, features, and debugging.
Create Playwright tests that check user journeys in a browser. The skill organizes reusable page interactions, configures test environments, investigates intermittent failures using execution traces, and supports API mocking, visual comparisons, and automated pipeline integration.
Build and maintain machine learning apps on Hugging Face Spaces. Choose a suitable app framework and computing hardware, load models, deploy files, and check logs and outputs to diagnose build or runtime problems.
Test polly's coordination of coding agents using a scripted mock language model and live command-line tools. Check startup, task dispatch, safety rules, parallel delegation, and whether real agents make appropriate orchestration decisions.
Plan a browser game's architecture before coding. Choose an engine, separate game rules from rendering, and define input, assets, menus, saves, debugging, and performance checks within a shared design.
Build browser-based 3D games with Three.js, TypeScript, and Vite. Structure scene and camera control, model loading, physics, and interface overlays separately from game rules, with guidance for WebGL rendering diagnostics and performance.
Review Python, JavaScript/TypeScript, and Go code against security guidance for the languages and frameworks in use. Identify important weaknesses, suggest improvements, and produce prioritized reports with code references, or apply secure defaults when writing new code.
Create specialized assistants for Cursor that work with their own context, separate from the main conversation. Define responsibilities, instructions, output formats, model choices, and execution settings for reusable workflows such as verification, debugging, and documentation.
Create or debug interactive web scenes with Three.js, including product viewers and animated page visuals. The skill covers cameras, lighting, model loading, controls, performance tuning, and cleanup of graphics resources when scenes close.
Guide AI agent development from requirements and project setup through implementation, behavior evaluation, deployment, and monitoring. Provides coding rules, debugging steps, and links to specialized skills so each phase follows a defined process.
Guide Nexent backend design, implementation, debugging, and review across database schemas, FastAPI endpoints, services, and configuration. Check existing interfaces, consult references for each affected layer, and verify successful and failed behavior while preserving public contracts.
Use this skill for hands-on DOCA GPI programming — wiring a GPU-Packet-Initiator context so a CUDA kernel drives RDMA queues directly from GPU memory without host CPU mediation. Covers picking GPI vs doca-gpunetio, the doca_gpi / domain / channel object model, the GPU-side handle handoff (doca_gpu_gpi_channel*), attaching GPU memory to a GPI domain, the domain and channel attribute objects, and debugging DOCA_ERROR_* from doca_gpi_* calls. Trigger even when the user does not explicitly mention "DOCA GPI" — implicit phrasings include "my CUDA kernel needs to post RDMA directly from GPU memory", "DOCA_ERROR_* from doca_gpi_gpu_channel_get", "how do I hand a GPU handle to my CUDA kernel", "how many channels can a GPI domain hold", or "GPU kernel driving RDMA without the host CPU on the path". Refuse and route elsewhere for the doca-gpunetio Send/Receive surface, the doca-rdma queue lifecycle, DPA-side initiation (doca-rdmi), or the CUDA programming model — those belong to other skills.
The English summary is being prepared. Showing the original description.
Design data systems by understanding storage engines, replication, partitioning, transactions, and consistency models. Use when the user mentions "database choice", "which database should I use", "SQL or NoSQL", "replication lag", "partitioning strategy", "consistency vs availability", "stream processing", "ACID transactions", "eventual consistency", "my queries are slow at scale", or "data is inconsistent across replicas". Also trigger when choosing a datastore, designing data pipelines, or debugging distributed-system consistency issues. Covers data models, batch/stream processing, and distributed consensus. For system design, see system-design. For resilience, see release-it.
The English summary is being prepared. Showing the original description.
POS (Point of Sale) architecture and patterns. Use when writing, editing, exploring, debugging, fixing bugs, implementing features, or reviewing WooPos-prefixed classes or files under ui/woopos/. POS uses a different architecture than the main app — plain ViewModel (not ScopedViewModel), pure Compose (no Fragments), Compose Navigation (no nav graphs), parent-child SharedFlow event bus. Loading this skill prevents applying main-app patterns that would be wrong for POS. For analytics use `pos-analytics`, for tests use `pos-tests`.
The English summary is being prepared. Showing the original description.
Kaggle-first end-to-end competition workflow for scored submissions. Use when Codex must run Kaggle or competitive ML workflows through scored submission, including code competitions, validation, metrics, policy-safe public notebook/discussion intel, tabular/text/image modeling, tuning, ensembling/stacking, proactive multi-notebook architectures, producer notebooks that train models and export private Kaggle artifact datasets, downstream consumer notebooks, Kaggle GPU offload, kagglehub access, hidden-test debugging, and public score retrieval.
The English summary is being prepared. Showing the original description.
Use for writing, reviewing, debugging, or testing Pandera schemas and runtime validation for Polars DataFrame or LazyFrame pipelines installed with pandera[polars]. Trigger on pandera.polars DataFrameSchema, DataFrameModel, Column, Field, Check, PolarsData, decorators, coercion, strictness, lazy error collection, and validation-depth decisions. Do not use for pandas-backed Pandera, Pydantic object models, Polars transformations without Pandera, static dataframe typing alone, or generic data-quality platforms.
The English summary is being prepared. Showing the original description.
Kaggle-first end-to-end competition workflow for scored submissions. Use when Codex must run Kaggle or competitive ML workflows through scored submission, including code competitions, validation, metrics, policy-safe public notebook/discussion intel, tabular/text/image modeling, tuning, ensembling/stacking, proactive multi-notebook architectures, producer notebooks that train models and export private Kaggle artifact datasets, downstream consumer notebooks, Kaggle GPU offload, kagglehub access, hidden-test debugging, and public score retrieval.
The English summary is being prepared. Showing the original description.
Use when writing or debugging prompts for ByteDance Seedance video models (Seedance 2.5, 2.0, 2.0 Mini, 1.5 Pro, 1.0) on Dreamina, Jimeng AI, Doubao, BytePlus ModelArk or ComfyUI, when a generated video drifts off the reference face, grows unwanted subtitles or watermarks, duplicates a character, jumps at an extension seam, or when planning multi-shot or multi-reference video briefs.
The English summary is being prepared. Showing the original description.
Developer reference for building applications, tools, and agents on the Pi Durable agent harness (@earendil-works/pi-durable) and the pi-ai model layer. Use when writing or debugging code that imports @earendil-works/pi-durable or pi-ai: assembling a Harness over memory/SQLite/JSONL/Durable Object storage, submitting input and waiting for answers, defining tools with TypeBox schemas and structured output, calling tools from tools, hooks and guards, durable tasks with checkpoints and ownership, subagents and child tasks, documents and migrations, code mode, tool search, per-conversation environments, compaction, recovery after a crash, storage and environment conformance tests, or choosing models and providers. Also use when a durable/agent-runtime behavior question comes up: what survives a restart, why an input stayed queued, why a task is blocked or orphaned, how abort cascades, or what the Pico5 specification requires.
The English summary is being prepared. Showing the original description.
Use when EZ_math_model coder or packaging fails repeatedly with the same error and needs root-cause analysis instead of another blind retry.
The English summary is being prepared. Showing the original description.
Use when the user is debugging a bug, tracing an error, or asking why something fails. Examples: "Why is X failing?", "Where does this error come from?", "Trace this bug"
The English summary is being prepared. Showing the original description.
Design, configure, debug, and optimize OpenClaw AI agent deployments. Master guide for gateway configuration, openclaw.json settings, model routing and fallback chains, skills development and publishing, cron job scheduling, memory systems (Qdrant, Neo4j, SQLite), Docker infrastructure, and Tailscale VPN networking. Includes config analyzer that audits your openclaw.json and suggests improvements, plus health checker that validates all OpenClaw subsystems. Built for AI agents — Python stdlib only, no dependencies. Use for OpenClaw setup, gateway debugging, skill building, cron management, model optimization, cost reduction, and infrastructure troubleshooting.
The English summary is being prepared. Showing the original description.