op1

Interview the user relentlessly about a plan or design until branch-level decisions are resolved for execution.

Access Figma designs, extract design systems, and retrieve component specifications. Use when implementing UI from Figma mockups, extracting design tokens, or analyzing design files.

Enforce cost-aware MCP usage. Use when a task might trigger heavy external tools, web search, or broad context expansion. Prevents token burn by ensuring MCPs are only used when local context is insufficient.

Navigate the Warmplane mcp0 facade efficiently. Use when the active config exposes provider capabilities through `mcp0_*` tools and you need to discover or call provider tools without brute-force describing large capability sets. Trigger on requests involving mcp0, Warmplane, or provider work through the facade such as Linear, Notion, Figma, New Relic, Context7, grep.app, or Storybook tools.

Use this when the user needs to control Chrome, navigate to a page, inspect a tab, click or fill elements, take screenshots, or automate a browser flow with aeroxy/chrome-devtools-cli.

Guidelines for creating and managing implementation plans with citations

Apply these conventions during planning and implementation from the start (not only after review feedback). Use for frontend and backend changes when deciding whether to add layers, where to place formatting/validation, how much typing to keep, how to place helper files (.dto.ts/.type.ts/.constant.ts/.util.ts), and how to match existing module patterns during /plan and /work.

ULTRAWORK MODE for high-stakes execution. Use for plan-driven work that needs parallel delegation, strict verification, and automatic follow-through. Prefer `/work` or an explicit ultrawork request.

Opt-in workflow for resumable multi-hour or multi-day execution using durable local state, append-only logs, and explicit pause/resume controls.

Run after making React changes to catch issues early. Use when reviewing code, finishing a feature, or fixing bugs in a React project.

Explore, build, and refine UI.

Context-first investigation mode. Use for debugging, system understanding, and any task where conclusions should follow grounded internal and external analysis.

Context and token management for long sessions. Load for complex work, heavy delegation, compaction safety, or when context bloat starts hurting signal quality.

Find documentation using llms.txt standard, context7.com, and repomix. Best practices for library research.

Exhaustive search mode. Use when you need broad codebase or documentation coverage, multiple parallel searches, and a synthesized inventory of findings.

Reusable simplification policy for preferring the current implementation, deleting compatibility glue, and collapsing legacy branches unless an explicit contract still requires them.

4-phase debugging framework preventing "guess-and-check" fixing. Load when debugging fails or for complex bug investigation.

Git operations mastery. Atomic commits, rebase/squash, history search (blame, bisect, log -S). Triggers on 'commit', 'rebase', 'squash', 'who wrote', 'when was X added', 'find the commit that'.

MANDATORY before creative or design-heavy work. Use to clarify intent, compare approaches, and shape implementation-ready designs one question at a time.

Authoritative code review rubric for changed files. Covers 4 review layers, severity, confidence, merge recommendation, and concrete file:line findings.

Forces agents to run verification commands and show evidence before claiming completion. Prevents "it should work" without proof.

Internal logic and data flow philosophy (The 5 Laws of Elegant Defense). Apply to backend, hooks, state management, and any code where functionality matters.

Terminal orchestration via tmux. Use for long-running processes, server management, parallel terminal sessions, and background task monitoring.

Frontend implementation and review guardrails for fail-closed API boundaries, schema validation, UX parity, async UI-state isolation, and deprecation hygiene. Use for FE coding and review across products.

Manage Lokalise translation keys and sync JSON locales for this monorepo. Use when adding new i18n keys, checking key existence in Lokalise, creating/updating keys, and pulling translation JSON files into packages/shared/translations/src/locales.

Respond to GitHub PR review comments (Copilot, human, or bot). Covers listing reviews, reading comments, posting replies, fixing code, and amending commits. Triggers on 'review comments', 'respond to review', 'copilot review', 'PR comments', 'reply to comments'.

Modern backend patterns, API design, OWASP security checklist, testing pyramid. For Node.js, Python, Go backends.

PostgreSQL and MongoDB patterns - queries, indexing, performance optimization, migrations.

Visual & UI philosophy (The 5 Pillars of Intentional UI). Apply to styling, layout, colors, typography, animations. Avoid "AI slop" aesthetics.

HTTP client patterns for axios, ky, and native fetch. Detect which client the project uses and follow its patterns. Use for API calls, request/response interceptors, error handling, retries, and TypeScript integration.

Remove AI writing patterns from text. Based on Wikipedia's AI Cleanup research with 24 pattern categories including inflated symbolism, promotional language, AI vocabulary, and formatting issues.

Guide for creating MCP (Model Context Protocol) servers in TypeScript or Python. Use when building MCP servers to integrate external APIs or services.

Comprehensive NestJS framework mastery with Drizzle ORM, TypeORM, Prisma, BullMQ queues, and 40 best practice rules. Use for NestJS applications, APIs, authentication, database integration, queue processing, microservices, testing, and architectural decisions.

Monitor applications, investigate performance issues, and analyze observability data in New Relic. Use when the user needs APM metrics, error tracking, infrastructure monitoring, or incident analysis.

Browser automation via Playwright MCP. Use for verification, browsing, web scraping, testing, screenshots, and all browser interactions.

React and Next.js performance optimization from Vercel Engineering. 45 rules across 8 categories covering async waterfalls, bundle size, server-side performance, re-renders, and rendering optimization. Use when writing, reviewing, or optimizing React/Next.js code.

System architecture design for scalable, maintainable systems. Includes architecture diagrams, design patterns, tech stack decisions, and dependency analysis.

Guide for creating effective skills that extend agent capabilities. Use when creating new skills or updating existing ones with specialized knowledge, workflows, or tool integrations.

Comprehensive Terraform/OpenTofu mastery covering testing (native 1.6+, Terratest), modules, CI/CD, security scanning (trivy, checkov), and production patterns. Use for IaC architecture, module development, testing strategy, state management, and infrastructure decisions.

Schema validation patterns for Zod and class-validator in NestJS. Detect which library the project uses and follow its patterns. Use for DTO validation, request/response validation, custom validators, and TypeScript type inference.