All Skills

Applies Rust rules for ownership, modules, errors, unsafe code, concurrency, tests, and tooling.

Applies safe ESPHome rules for YAML, lambdas, sensors, switches, relays, boot behavior, fail-safes, and validation.

Strict test-first gate only when explicitly requested or required by repository policy. Do not use for ordinary practical testing.

Improves measured performance while preserving correctness, reliability, and maintainability.

Builds and validates Russian Obsidian repository wikis, architecture docs, and code/docs/ADR drift reports.

Implements Figma frames, selections, components, or screens in code while preserving assets, variants, tokens, and visual fidelity.

Selects session runtime capabilities when tools or external evidence are required. Do not use for ordinary local repository work.

Establishes visual direction from Figma, screenshots, generated references, evidence, or an existing design system before implementation.

Reviews lockfiles, installers, provenance, CI permissions, dependency confusion, artifacts, registries, and publishing.

Use only when the request concerns standards, responsibility, complexity, smells, naming, refactoring catalogs, or task-plan rules; not routine work.

Decomposes large modules, god files, classes, or packages in reviewable steps while preserving public APIs. Not for routine refactors.

Reviews secrets, config defaults, environment variables, logs, credentials, CI secrets, and unsafe settings.

Reviews pull requests and local diffs against Google's eng-practices code review guide (design, complexity, tests, naming, comments, style, consistency, documentation, every-line, context) and emits a ranked, principle-tagged finding list with rich local diff rendering. Optional follow-up step posts selected findings as a GitHub PR review. Invoke when the user says "googly-eyes this PR", "review by the canon", "principled code review", "Google-style review", "review this CL", or pastes a PR URL and asks for a code review. 👀

补充上下文，用更简单、明确的中文重新解释。

当用户提出一个想法或诉求时，通过对话挖掘真实意图、明确边界，提炼成清晰、可执行的概念。

当用户明确要求“解决当前 merge / rebase / cherry-pick 冲突、提交代码、创建 commit、把改动分批提交并 push、推送当前分支”时使用。进行中的冲突要先识别 Git 操作和双方原始意图，逐 hunk 做语义合并并运行项目检查，再经用户确认继续操作；普通改动要检查敏感信息与异常差异，拆成可运行、可审查、可回滚的原子提交后确认提交和推送。不得自动发起或中止 merge / rebase / cherry-pick，不得强推或随意改写已发布历史。不要用于普通改动尚未完成或验证失败、只想查看 Git 状态，或要创建 PR 的场景。

Use for creating, reviewing, debugging, or testing dbt SQL projects involving models, sources, ref, tests, macros, materializations, incremental models, snapshots, seeds, selectors, compiled SQL, or dbt artifacts. Do not use for generic SQL with no dbt graph, warehouse administration, semantic-layer product configuration, or orchestration outside dbt.

Use for writing, reviewing, debugging, testing, or diagnosing Bayesian regression and hierarchical models built with Bambi formulas, Model, Family/Likelihood/Link, Prior, fit, prior predictive, and predict. Trigger on common versus group-specific terms, categorical coding, family/link choice, automatic prior scaling, missing rows, PyMC backend settings, and InferenceData predictions. Do not use for hand-built PyMC graphs, NumPyro programs, ArviZ-only analysis of existing draws, frequentist statsmodels formulas, or generic pandas work.

---

Use for writing, reviewing, debugging, testing, or optimizing Python code that embeds DuckDB, executes analytical SQL, manages DuckDB connections and transactions, builds DuckDB relations, queries Parquet/CSV/Arrow/pandas/Polars inputs, or exports query results. Trigger on connection scope, parameters, replacement scans, materialization, concurrency, extensions, and query plans. Do not use for generic SQL with another engine, DuckDB CLI-only work, server-database administration, dbt-only projects, or dataframe work that does not call DuckDB.

Write, review, debug, test, and optimize Python Polars code with version-grounded object types, schemas, and execution boundaries.

Write, review, debug, or test Python asyncio code involving coroutines, Tasks, TaskGroups, cancellation, timeouts, queues, synchronization, and blocking-call boundaries.

Write, review, debug, test, or optimize Python code using PyArrow arrays, schemas, tables, compute kernels, datasets, Parquet, and Arrow IPC.

Use for writing, reviewing, debugging, or testing bounded retry policies in Python with Tenacity, including retry predicates, stop and wait strategies, jitter, exception propagation, callbacks, Retrying, and AsyncRetrying. Do not use for generic loops, scheduled jobs, domain polling without a retryable operation, or operations whose side effects are not safe to repeat.

Use for writing, reviewing, debugging, or testing static Python visualization with Matplotlib, including Figure, Axes, Axis, Artist, transforms, layouts, dates, categorical scales, annotations, color normalization, image export, and headless rendering. Do not use for Plotly interactivity, Altair/Vega-Lite specifications, dashboard state, or analysis without a Matplotlib artifact.

Use for designing, creating, reviewing, or debugging Model Context Protocol server configuration for GitHub Copilot in Visual Studio Code, including .vscode/mcp.json, transports, inputs, trust, sandboxing, and tool scope. Do not use for implementing an MCP server, ordinary HTTP clients, or silently installing or starting external software.

Use when writing, reviewing, debugging, or testing PowerShell `.ps1` scripts and advanced functions, including parameters, pipeline objects, streams, native commands, filesystem safety, PowerShell 7 portability, Pester, and PSScriptAnalyzer. Do not use merely to run an existing script, author Bash or batch files, package a module or DSC resource, or perform live system or tenant administration.

Guides the choice between DuckDB and Polars for each stage of an analytical pipeline, covering Arrow transfer, lazy versus eager execution, registration lifetime, schema conversion, and result ownership.

Use for creating, splitting, reviewing, or debugging GitHub Copilot custom instructions in Visual Studio Code, including .github/copilot-instructions.md, .github/instructions/*.instructions.md, and AGENTS.md. Do not use for slash commands, Agent Skills, custom agents, hooks, or ordinary project documentation.

Use for designing, auditing, or refactoring a GitHub Copilot customization system in Visual Studio Code across instructions, prompt files, Agent Skills, custom agents, hooks, MCP servers, and plugins. Do not use merely to author one already-selected artifact or configure unrelated VS Code settings.

Follow an open PR with the client-neutral cenci supervisor until it merges or closes.

Claude Code-only: verify the cenci launcher and build the sandbox container image. Codex users receive the launcher through the cenci installer.

Git worktree patterns for isolated parallel development. Use when creating feature branches, managing git worktrees, isolating feature work, parallel development, setting up a worktree, listing worktrees, worktree naming conventions, or cleaning up worktrees.

.NET backend conventions, patterns, and test infrastructure. Use when working with C# files, .csproj projects, ASP.NET APIs, xUnit tests, FluentAssertions, Entity Framework, WebApplicationFactory, dotnet CLI, NuGet packages, .NET dependency injection, middleware, or controller patterns.

Establish and verify exclusive ownership of a GitHub issue before refine, design, or implementation work begins. Use from ticket-mode workflows that must claim an unassigned issue for the active GitHub CLI user without replacing existing assignees.

Guide for continuous improvement, error proofing, and standardization. Use this skill when the user wants to improve code quality, refactor, or discuss process improvements.