korallis

Set up and maintain continuous integration and continuous deployment pipelines with GitHub Actions, GitLab CI, Jenkins, or similar tools to automate testing, building, and deployment. Use when configuring automated builds, setting up test automation,…

Build accessible web applications following WCAG 2.1/2.2 guidelines with proper semantic HTML, ARIA attributes, keyboard navigation, screen reader support, and inclusive design. Use when implementing ARIA labels and roles, ensuring keyboard navigation,…

Design clean, scalable, and maintainable REST and GraphQL APIs following industry best practices. Use when designing public or internal APIs, planning endpoint structures, defining request/response contracts, establishing versioning strategies, implementing…

Create interactive, self-contained artifacts like mini-apps, visualizations, calculators, and tools that can be embedded or used standalone. Use when building interactive calculators, creating data visualizations, making mini web apps, building embeddable…

Architect scalable backend services using layered architecture, dependency injection, middleware patterns, service classes, and separation of concerns. Use when building API services, implementing business logic layers, creating service classes, setting up…

Facilitate ideation, explore alternatives, and refine concepts through structured brainstorming techniques before implementation. Use when starting new projects, designing features, exploring solutions, evaluating tradeoffs, challenging assumptions,…

Create and maintain comprehensive brand guidelines including visual identity, tone of voice, messaging, color palettes, typography, and usage rules for consistent brand expression. Use when establishing brand identity, creating style guides, defining visual…

Create visual designs, graphics, and layouts using HTML Canvas API or design tools for data visualization, graphics, animations, and interactive visual content. Use when creating data visualizations, building interactive graphics, implementing drawing tools,…

Clean Architecture. Enforce the Dependency Rule. Use when adding a file, import, module, package, layer, dependency, framework, database, HTTP client, or use case, or when deciding where code belongs.

Clean Code. Enforce Martin's naming, function, comment, error, test, and smell rules on every change. Use when writing, editing, reviewing, or refactoring any source file, or when the user mentions naming, functions, comments, duplication, Boy Scout, smells,…

Global TypeScript-first engineering router for Lee's projects. Use whenever planning, designing, building, implementing, refactoring, debugging, reviewing, or documenting software. Selects the smallest relevant specialist skills, local multi-agent workflows,…

Scaffolding. Enforce Dune-style capability folders, public module contracts, one data owner, and mechanical boundaries. Use when creating an app, module, package, workspace, monorepo, or folder tree, or when deciding where a new capability lives.

Architecture-first workflow for non-trivial or high-risk software changes. Use when a plan or implementation crosses modules, introduces domain concepts, changes contracts, or risks shallow layering. Grounds designs in callers, data structures, signatures,…

Local multi-agent design or implementation competition. Use when several materially different solutions are plausible and selection quality matters. Creates isolated candidates, judges them against shared criteria, and lets the lead agent synthesize and…

Change-impact and completion audit for wide, risky, cross-cutting, or merge-critical work. Use before declaring such work complete to inspect callers and contracts beyond the diff, identify safety facts, exercise the real artifact, and report what is cleared…

Review the changes since a fixed point (commit, branch, tag, or merge-base) along two axes: Standards (does the code follow this repo's documented coding standards?) and Spec (does the code match what the originating issue/spec asked for?). Runs both reviews…

Plan and start isolated parallel module implementation tracks from specs/MODULES.md and per-module parallel.yaml declarations.

Implement a fully-specced module via dual-harness execution — Claude orchestrates, Codex CLI executes inside a live tmux pane in your attached session. Use when ready to write code for a module that has approved SPEC.md and CLAUDE.md files. Triggers on…

Initialise a new software project with full spec-first workflow — research, architecture, module specs, and roadmap. Use this whenever someone says 'build me a...', 'I want to create...', 'new project', 'start a project', or describes a product idea. Also use…

Generate or regenerate the master architecture document (specs/MASTER_BLUEPRINT.md) — the single source of truth for tech stack, data model, API patterns, auth, UI conventions, and module relationships. Use when starting architecture, changing stack…

End-of-session review — captures what was done, learnings, mistakes, and patterns, then updates LEARNINGS.md and CLAUDE.md. Run this after completing any task, at the end of a working session, or when switching to a different area of the project. This is what…

Independent security review of pending changes via an isolated Agent context — fresh subagent reads the diff and a security checklist with no implementation bias. Use after a module is implemented and before merge, especially for changes that touch auth, data…

Create or update a detailed module specification — data model, API endpoints, UI screens, business logic, and acceptance criteria. Use when planning a new module, refining requirements before implementation, or updating specs after implementation revealed…

Run a comprehensive test pass across the project — unit tests, type checking, linting, and visual verification with browser automation. Use when someone says 'run the tests', 'check everything works', 'test the project', or before deploying to verify quality.

Agent-only reference for firstmate harness operations. Use before spawning or recovering a crewmate or secondmate, handling a trust dialog, sending a harness-specific skill invocation, interrupting or exiting an agent, resuming an exited agent, or verifying a…

Generate a "pick up where I left off" status report from firstmate's live fleet state. Use when the captain invokes /bearings or asks for a bearings report, morning brief, status report, catch-up, "where did I leave off", or "what's in the works". Reads live…

Agent-only reference for persistent secondmate setup and retirement. Use when creating, seeding, validating, launching, recovering, handing backlog to, pushing inherited config into, or retiring a secondmate home, or when editing data/secondmates.md. Covers…

Self-update a running firstmate and its secondmates to the latest from origin. Use when the captain invokes /updatefirstmate (e.g. "/updatefirstmate", "update firstmate", "pull the latest firstmate"). Fast-forwards this firstmate repo's default branch and…

Agent-only playbook for coordinating visible Codex Desktop threads alongside Firstmate without pretending they are a selectable shell backend. Use before creating, reading, steering, archiving, debugging, or reviewing a Codex App visible thread for Firstmate…

Agent-only playbook for handling X mode mentions and follow-ups. Use on an "x-mention <request_id>" check wake to read the stashed mention, classify it, act autonomously on eligible requests, reply or dismiss, and link spawned work. Also use on an…

Enter away-mode supervision. Use when the user invokes /afk (e.g. "/afk", "/afk back in an hour", "going afk"). Sets a durable away-mode flag so the sub-supervisor daemon can self-handle routine wakes and escalate only captain-relevant events as one batched…

Agent-only reference for changing firstmate's shared, tracked material per AGENTS.md section 1. Use before editing any of that material, whether working as firstmate directly or as a crewmate briefed on a firstmate-repo task. Covers the knowledge-placement…

Implement a fully-specced module via dual-harness execution — Claude orchestrates, Codex CLI executes inside a live tmux pane in your attached session. Use when ready to write code for a module that has approved SPEC.md and CLAUDE.md files. Triggers on…

Generate or regenerate the master architecture document (specs/MASTER_BLUEPRINT.md) — the single source of truth for tech stack, data model, API patterns, auth, UI conventions, and module relationships. Use when starting architecture, changing stack…

Initialise a new software project with full spec-first workflow — research, architecture, module specs, and roadmap. Use this whenever someone says 'build me a...', 'I want to create...', 'new project', 'start a project', or describes a product idea. Also use…

Implement a module that already has an approved spec — reads all specs, enters plan mode, then builds incrementally with tests and commits. Use when ready to write code for a specific module, or when someone says 'build the X module', 'implement X', or 'start…

Conduct deep, structured research on any domain, technology, competitor, regulation, or technical topic for this project. Use whenever the team needs to understand something before making architectural or product decisions — market landscapes, compliance…

End-of-session review — captures what was done, learnings, mistakes, and patterns, then updates LEARNINGS.md and CLAUDE.md. Run this after completing any task, at the end of a working session, or when switching to a different area of the project. This is what…

Agent-only reference for firstmate harness operations. Use before spawning or recovering a crewmate or secondmate, handling a trust dialog, sending a harness-specific skill invocation, interrupting or exiting an agent, resuming an exited agent, or verifying a…

Validate your code changes through the no-mistakes pipeline - automated code review, tests, lint, docs, push, PR, and CI - before they reach the configured push target. Use when the user asks to run no-mistakes, gate or ship or validate their changes, push…

Bootstrap agents for iOS, iPadOS, macOS, Swift, SwiftUI, SwiftData/Core Data, Swift Testing, Xcode build/test/debug, Simulator, App Intents, or XcodeBuildMCP work. Use before building, fixing, refactoring, QAing, or setting up Apple-platform repos, and when…

Vercel Functions expert guidance — Serverless Functions, Edge Functions, Fluid Compute, streaming, Cron Jobs, and runtime configuration. Use when configuring, debugging, or optimizing server-side code running on Vercel.

Enter away-mode supervision (agents copy — should lose to the .claude copy).

Only ships under .agents/skills, not user-invocable.

Enter away-mode supervision (claude copy — should win over the .agents copy).

Sweep the current session for uncaptured durable knowledge and file it to disk before a context reset.

Agent-only reference for firstmate harness operations. Use before spawning or recovering a crewmate or secondmate, handling a trust dialog, sending a harness-specific skill invocation, interrupting or exiting an agent, resuming an exited agent, or verifying a…

Enter away-mode supervision. Use when the user invokes /afk (e.g. "/afk", "/afk back in an hour", "going afk"). Sets a durable away-mode flag so the sub-supervisor daemon can self-handle routine wakes and escalate only captain-relevant events as one batched…

Agent-only reference for persistent secondmate setup and retirement. Use when creating, seeding, validating, launching, recovering, handing backlog to, pushing inherited config into, or retiring a secondmate home, or when editing data/secondmates.md. Covers…

Self-update a running firstmate and its secondmates to the latest from origin. Use when the captain invokes /updatefirstmate (e.g. "/updatefirstmate", "update firstmate", "pull the latest firstmate"). Fast-forwards this firstmate repo's default branch and…

Full-stack feature development for Complete Care — database, server actions, API routes, and UI. MUST invoke the frontend-design skill for all UI work.

Handles project scaffolding, database schema, auth, middleware, and backend infrastructure for the Complete Care platform.

Implements Python code for the VibeGuard security scanner using TDD — writes tests first, then implements CLI, config, stack detection, rules, engines, reports, auto-fix, and LLM integration.

Creates intentionally vulnerable and clean test fixture apps in tests/fixtures/ for validating VibeGuard scanner rules across all supported stacks.

Orchestrates significant coding work through kor-pi's fusion-native scout, planner, gated builder, dual reviewers, and documenter. Use for multi-file features, risky refactors, audits, architectural decisions, or work needing cross-provider validation.