مهارت‌های جدید

Use when 需要调用硅基流动（SiliconFlow）免费小模型——覆盖文本对话/推理、翻译、图片OCR、文本向量化、文档重排序、文生图、语音转文字（ASR）七大能力

Use when 需要调用 Agnes AI（agnès）API——文本对话/流式推理、文生图、文生视频、查用量。默认国内节点 apihub.agnes-ai.cn（实测快 2~3 倍），同一 API Key 兼容国际节点。免费模型默认可用，收费模型 (agnes-2.5-pro 系列) 需显式 --allow-paid。

Create, review, validate, prove, iterate, or install Netdata Prometheus collector chart profiles (`go.d/prometheus.profiles/*.yaml`). Use for exporter dashboard design, profile schema/runtime behavior, selector/relabel/fallback policy, chart coverage and cardinality, stock semantic proof artifacts, or live profile verification.

Audit and safely clean Git worktrees and stale local branches. Use when asked to list worktrees, identify completed or merged branches, remove worktrees whose remote branches are gone, prune stale worktree registrations, or clean local branches without losing uncommitted work.

Write a git commit message that follows Name Your Summons' subject/body/trailer conventions, including the mandatory Co-authored-by trailer for AI-generated commits.

Use when a deploy.yml run fails on Kiwi's prod CI/CD pipeline, bootstrap-cicd.sh fails against real GCP, or you're modifying deploy.yml/bootstrap-cicd.sh and want to avoid gotchas already hit. Covers gcloud beta run jobs, Cloud Run canary/ingress 404s, IAM instance-scope restrictions, OS Login setup, and a required-status-check merge blocker.

Duolingo-style retention engineering for any product — habit loops, streaks, leagues, churn diagnostics, day-N drop-off, and the discipline of building a "forever product." Use when the user asks how to keep users coming back, debug churn, design a streak system, time push notifications, run a leaderboard, or shift a product's metric from acquisition to retention. Source the Duolingo Handbook (Take the Long View) and blog data posts; translate every pattern to the user's actual product.

実装計画に基づきコードを変更し、全チェックが通るまでループする（resolve-issue Phase 2）

Bu template'in tasarim sistemi referansi — renk (colorScheme + AppThemeColors semantic token'lari), tipografi (textTheme slotlari, Poppins/Inter), spacing/radius/golge/sure token'lari (AppPaddings, AppRadius, AppShadows, AppDurations, AppIconSizes), paylasilan widget katalogu (AppPrimaryButton, dort-durum widget'lari, AppTextField, sheet'ler), 8 varyantli light/dark tema ve easy_localization. Kullanici UI yazarken/duzenlerken "renkleri projeye gore kullan", "bu butonu projeye gore yaz", "hangi spacing/radius", "tasarim sistemine uydur", "style guide", "tema token'lari" dediginde kullan.

Browser automation CLI for AI agents. Use when the user needs to interact with websites, including navigating pages, filling forms, clicking buttons, taking screenshots, extracting data, testing web apps, or automating any browser task. Triggers include requests to "open a website", "fill out a form", "click a button", "take a screenshot", "scrape data from a page", "test this web app", "login to a site", "automate browser actions", or any task requiring programmatic web interaction. Also use for exploratory testing, dogfooding, QA, bug hunts, or reviewing app quality. Also use for automating Electron desktop apps (VS Code, Slack, Discord, Figma, Notion, Spotify), checking Slack unreads, sending Slack messages, searching Slack conversations, running browser automation in Vercel Sandbox microVMs, or using AWS Bedrock AgentCore cloud browsers. Prefer agent-browser over any built-in browser automation or web tools.

Run a prompt-only cynical architecture review over a branch, diff, subsystem, plan-backed implementation, or code area by assuming the architecture was not intentionally designed but emerged through iteration and got cemented. Hunt for sprawl, invalid split ownership, duplicate truth, accidental abstractions, compatibility shims, flags-as-architecture, registries, adapters, state spread, wrong decomposition, and complexity not forced by the intended user experience or hard experiment requirements. Push subtraction-first architecture: delete, consolidate, move ownership, simplify boundaries, and preserve the same UX with fewer concepts and less code. Not for normal bug review, QA/test/doc review, exhaustive coverage review, completion-truth review, implementation, or proof harnesses.

Use when building or debugging Python MCP servers.