Creates specs before coding. Use when starting a new project, feature, or significant change and no specification exists yet. Use when drafting a PRD or requirements document with objectives and scope, or when requirements are unclear, ambiguous, or only exist as a vague idea. Use when a single requirement spans several independently testable capabilities and needs decomposing into a capability map of modules before specifying.
Encontra oportunidades de "deepening" no codebase — refactors que transformam modulos
shallow (interface complexa, implementacao simples) em deep (interface simples,
implementacao rica). Foco em testabilidade e AI-navigability. Use quando usuario
quiser melhorar arquitetura, encontrar oportunidades de refactor, consolidar modulos
acoplados, ou preparar codebase para trabalho de agente.
Trigger em: "deepening", "deep module", "shallow module", "refactor architecture",
"improve architecture", "consolidate modules", "agent-friendly codebase",
"AI-navigable", "module depth".
Turn a goal into an approved, phased implementation plan where every phase ends at a real test checkpoint. Use before implementing anything non-trivial, or when a task has several viable approaches.
Configure Vercel CI/CD with GitHub Actions, preview deployments, and automated testing.
Use when setting up automated deployments, configuring preview bots,
or integrating Vercel into your CI pipeline.
Trigger with phrases like "vercel CI", "vercel GitHub Actions",
"vercel automated deploy", "CI vercel", "vercel pipeline".
Use to execute a dev-flow plan task by task. This is step 5 of the dev-flow workflow — work the plan's tasks with TDD (failing test → implement → green → commit per task), updating the state file as you go. The Implementer dispatches per the state file's Roles block lane: inline (runs in this session), session, claude:<name>, or cli:<tool>/<model> (external CLI via thin-driver). Can be invoked standalone if a plan and state file already exist.
Think and work like an expert Catalysis Scientist. Use when a task calls for Catalysis Scientist judgment. Reasons from active-site structure, turnover frequency, selectivity, and the Sabatier principle through CO/H2 chemisorption site-counting, Weisz-Prater and Mears transport checks, Langmuir-Hinshelwood/Mars-van Krevelen kinetics, and operando DRIFTS/XAS while treating diffusion-limited apparent rates, DRIFTS spectator species, mercury-test false positives for leached Pd, sintering, coking, and selectivity collapse at high conversion as first-class failure modes.
Prepare, run, or diagnose connected NetSuite sandbox validation and render-parity QA for this PDF engine and template pack. Use for validate, deploy, render, or save-parity work; do not use for ordinary local implementation or unit tests.
Build, run, and drive ircx. Use when asked to start ircx, run its tests, build it, take a screenshot of its UI, click through the app, or check that a frontend change works in the running client.
Measure observable complexity, dependency, test, build, and database effects of changes in Ruby, Rails, React, TypeScript, and PostgreSQL without fabricating savings.
Workflow 1: Full idea discovery pipeline. Orchestrates research-lit → idea-creator → novelty-check → research-review to go from a broad research direction to validated, pilot-tested ideas. Use when user says "找idea全流程", "idea discovery pipeline", "从零开始找方向", or wants the complete idea exploration workflow.
Eng manager-mode plan review. Lock in the execution plan — architecture,
data flow, diagrams, edge cases, test coverage, performance. Walks through
issues interactively with opinionated recommendations. Use when asked to
"review the architecture", "engineering review", or "lock in the plan".
Proactively suggest when the user has a plan or design doc and is about to
start coding — to catch architecture issues before implementation.
Run a premortem on any plan, launch, product, hire, strategy, or decision. Assumes it already failed and works backward to find every reason why. Produces a revised plan with blind spots exposed, likelihood/impact scores on each failure mode, and a concrete committed action. MANDATORY TRIGGERS: 'premortem this', 'premortem my', 'run a premortem', 'what could kill this', 'future-proof this', 'stress test this plan', 'what am i missing here', 'find the blind spots'. STRONG TRIGGERS: 'what could go wrong', 'am i missing anything', 'poke holes in this', 'where will this break', 'devil's advocate this'. Do NOT trigger on simple feedback requests, factual questions, or LLM Council requests. DO trigger when someone has a plan or commitment where the cost of being wrong is high.
Layered architecture with dependency inversion. Domain/Application/Presentation/
Infrastructure separation. For complex business logic, long-term maintainability,
testability. AVOID for simple CRUD apps or MVPs. Foundational pattern.
Use when AI-written code needs structural cleanup or when implementing a change under a guard against AI coding slop — duplicated reimplementations of existing helpers, try-import fallbacks, broad exception swallowing, speculative abstraction layers, _v2 copies, dead "for future use" code, narrating comments — including requests to "deslop", "remove AI slop", "humanize code", review an AI-generated PR beyond tests passing, or prevent these patterns in the current change.