Coding & Dev Tools

Delivers measurement-driven performance optimization for latency, throughput, memory, and tail behavior, with correctness preservation and regression guards.

Clarifies ambiguous or conflicting requests by researching first, then asking only judgment calls. Use when prompts say "grill me", ask hard questions, request relentless interrogation, pressure-test assumptions, clarify scope/requirements, define success criteria, or request system-design/optimization decisions.

Analyzes a local codebase to infer its actual architecture from code-first evidence, producing a dominant architecture label, coexisting patterns, drift, and repo-fit advice for downstream agents.

Generates a concise bullet-point summary of branch changes for a pull request description, grouping related commits by theme and verifying details against actual files.

Mines a codebase for evidence-backed improvement opportunities, forcing two escalation gates to surface breakthrough ideas, and outputs a ranked portfolio with a plan seed without implementing.

Reviews development plans for gaps, hidden assumptions, critical misses, and XP violations, then presents prioritized findings for confirmation.

Runs a UI redesign experiment by creating three git worktree branches for modern minimal, bold creative, and classic refined design directions, then implements and compares each approach.

Acceptance-test newly added features using browser automation, happy-path flows, screenshots, UX notes, and a written report.

Automates browser interactions for web testing, form filling, screenshots, and data extraction.

Refactors test files to improve behavioral coverage, clarity, structure, assertions, fixture setup, and mocking strategy, removing implementation-detail or low-value tests.

Refactors code by renaming variables, methods, classes, and types to improve clarity and intent, with automatic testing and commits after each rename.

Classifies code review findings to choose the smallest correct response, preventing one-patch-per-comment churn.

Refactors long methods or large code blocks by extracting small, focused helper methods with clear names and narrow responsibilities.

Define owned inductive invariants and enforce them at parse, construct, API, DB, lock, or transaction boundaries, with counterexample-driven verification.

Refactors Swift code toward declarative functional programming with immutability, value types, pure functions, transformations, and closure-based APIs.

Verifies that tests for already-implemented behavior are meaningful by temporarily breaking the production code and confirming the test fails for the expected reason.

Systematically explore an unfamiliar codebase and build a reusable architecture summary with file:line evidence, starting from documentation and tracing data flow.

Creates standardized git commits using the Conventional Commits specification by analyzing the diff, staging files, and generating appropriate commit messages.

Reviews all uncommitted changes in the current Git repository and produces a structured code review report with findings and a verdict.

Controls a live browser session on scraped pages to click, fill forms, navigate flows, and extract data using natural language prompts or code, including support for authenticated sessions via profiles.

Guides the creation, iterative improvement, and evaluation of agent skills, including drafting, testing, benchmarking, and optimizing descriptions.

小红书内容发现与分析技能。搜索笔记、浏览首页、查看详情、获取用户资料。 当用户要求搜索小红书、查看笔记详情、浏览首页、查看用户主页时触发。

Optimizes React and Next.js performance with 69 prioritized rules covering waterfalls, bundle size, server-side rendering, and re-rendering patterns.

Adds smooth, native-feeling animations to React apps using the View Transition API, covering shared elements, route changes, and list reordering with CSS-based control.

Search locally saved Codex session history stored in SQLite and session_index.jsonl, with filters for provider, time range, and archived status.

Drafts GitHub pull request titles, bodies, issue-closing footers, and maintainer ping comments, tailored to repo-specific commitlint or semantic PR conventions.

Ingests JavaScript-heavy external pages into StaticFlow as standalone local interactive mirrors backed by LanceDB, with bilingual article write-back and localized interactive locales.

Runs bottom-up type-chain validation across database, hooks, and UI layers, then fixes errors at the failing layer for project health checks before large changes.

Builds or debugs Anthropic Claude integrations using @anthropic-ai/sdk, covering messages, tool use, extended thinking, prompt caching, and streaming.

Build and manage Trigger.dev v4 background jobs: durable TypeScript tasks, scheduled/cron jobs, waits, Realtime, React hooks, and the Management API, with live docs fetched first.

Safely land one or many topic branches onto main via GitHub PR: inventory, commit leftovers, push, open/update PR, wait for green CI, merge or squash (repo norm), sync local main, then delete local/remote branches only after main contains the work.

Creates or modifies agent tools for OpenAI Agents SDK, covering search, analysis, API integration, and database query patterns with caching and metrics.

Migrates a Stitch screen to production React: fetches and caches wireframes, archives existing pages, converts to shadcn/ui components with semantic tokens, and documents the design system layers.

Builds or modifies a RAG pipeline with intent-based routing, vector store search, citation rendering, and book fidelity enforcement.

Regenerates a specific code layer (schemas, services, routes, hooks, ui) or all layers, then runs the corresponding validation checks.

Applies universal TypeScript, JavaScript, React, and Node.js quality conventions for this repo. Use when reviewing or writing TS/TSX, refactoring for clarity, or enforcing naming, immutability, and error-handling patterns.