Enforces semantic HTML, ARIA, keyboard navigation, and WCAG compliance. Audits and fixes accessibility across UI components. WHEN to use it - designing or implementing UI components, auditing existing interfaces, writing HTML markup, implementing complex interactive patterns, or addressing compliance requirements.
Designs or reviews API endpoints, request/response contracts, and versioning. Ensures consistency, clarity, and backward compatibility. WHEN to use it - designing new API endpoints, versioning existing APIs, reviewing API changes, documenting API behavior, or planning deprecation.
Designs system structure, data flows, and service boundaries. Prevents silent coupling and ensures systems evolve cleanly. WHEN to use it - when making design decisions affecting multiple apps/services, creating new content types or model changes, or evaluating whether a feature needs architectural planning.
Brand guide template defining voice, audience, visual identity, and content pillars. All marketing personas read this before producing output. WHEN to use it - when filling out brand identity for a new project, when marketing skills need brand guidance, or when reviewing brand consistency across output.
GitHub Actions CI/CD setup and configuration. Automates testing, building, and deployment with staging gates and approval workflows. WHEN to use it - setting up GitHub Actions workflows, configuring automated testing and deployment, managing secrets in CI/CD, or implementing deployment gates and approvals.
Infrastructure architecture, IaC, and cloud security posture. Designs scalable, resilient architectures using Infrastructure as Code. WHEN to use it - designing system architecture, setting up deployment pipelines, configuring cloud resources, diagnosing production outages, or implementing security controls at the…
Assesses confidence level and risk before acting on low-certainty tasks. Provides a structured scoring framework and escalation criteria. WHEN to use it - before starting any task, after completing implementation, when encountering unexpected behavior, when making architectural decisions, or when uncertain about…
Manages limited context window — decides what to load and what to release. Uses just-in-time loading, surgical reads, and summarize-and-release patterns. WHEN to use it - before starting any task, when context window is filling up, when navigating large codebases, or when working across multiple files.
Designs schemas, model relationships, and migration strategies. Ensures data integrity and performance through normalization, indexing, and constraints. WHEN to use it - designing new models/tables, planning database migrations, optimizing query performance, reviewing data relationships, or scaling database design.
Investigates bugs with systematic root cause analysis using hypothesis testing. Does not fix — diagnoses and hands off. WHEN to use it - bug reports in production or development, unexpected behavior or errors, intermittent failures, performance degradation, or after failed deployment.
Evaluates competing technical or product options using explicit criteria. Extracts assumptions, maps risk surfaces, and forces a documented decision. WHEN to use it - architecture decisions with long-term consequences, build vs buy decisions, deployment strategy changes, or any decision with irreversible or high-cost…
Implements backend, frontend, or fullstack tasks. Builds features with clarity, speed, and quality. WHEN to use it - for any implementation task, building views, components, forms, migrations, routing, and SEO configuration. Applies Logic → API → UI sequence.
Handles test, build, or deployment failures autonomously without human escalation. Analyzes root causes, implements fixes, and knows when to escalate. WHEN to use it - when tests fail, builds fail, deployments fail, unexpected errors occur, or when stuck on a problem.
Optimizes client-side performance for fast, responsive user experiences. Focuses on Core Web Vitals, bundle size, image optimization, and JavaScript performance. WHEN to use it - optimizing page load times, Core Web Vitals, bundle sizes, or client-side rendering performance.
Enforces commit message format, changelog discipline, and file organization. Ensures git history is useful and documentation is current. WHEN to use it - before committing code, when creating or updating documentation, when organizing project files, or during code reviews involving git history.
Code review for security vulnerabilities, bad boundaries, and convention drift. Reviews clarity, correctness, and safety while detecting unsanitized input/output, hardcoded secrets, and framework violations. WHEN to use it - after @dev completes implementation, when reviewing AI-generated code, before merging any…
Persists agent state, decisions, and execution plans across sessions using a Write-Ahead Log protocol. WHEN to use it - starting multi-session tasks, when context window is filling, before pausing work for review, or resuming work in a new chat session.
Ad copy, landing pages, UTM tracking, and A/B test design. Optimizes for measurable conversion outcomes. WHEN to use it - planning or creating paid ad campaigns, writing ad copy variants, designing conversion funnels, setting up UTM tracking, analyzing campaign performance, or creating landing page copy.
Identifies and fixes backend performance bottlenecks in databases, queries, and APIs. Measures first, optimizes for N+1 queries, missing indexes, and caching strategies. WHEN to use it - when performance issues are reported, before launching high-traffic features, during code review for inefficiencies, or quarterly…
Translates vague ideas into scoped, testable specs. Asks clarifying questions and marks in/out of scope explicitly. WHEN to use it - translating vague requirements into buildable briefs, breaking large features into shippable increments, defining acceptance criteria, scoping releases, or writing feature briefs for…
Philosophy: Every failure is a process failure, not a people failure. The goal is not to assign fault — it is to change the system so the failure cannot happen again.
Tests edge cases, regressions, and coverage gaps. Executes test suites and validates functional and integration behavior. WHEN to use it - after @guard passes before merging, validating edge cases on new features, or checking for regressions after refactoring.
★not rated 16 8d agoA51 tokens
At most 3 mods per repository are shown here, and a mod shipped inside a plugin is left to that plugin's page — the rest are on their repository pages: