sloth-skill-tree

Use when an approved Maycrest spec needs turning into a bite-sized TDD implementation plan a contractor could execute cold — before any code is touched on a client build.

Use when the Maycrest deliver pipeline executes an implementation plan of independent tasks in the current session — one fresh subagent per task, two-stage review each.

Use when a Maycrest build hits any bug, test failure, or unexpected behavior on the Expo/Supabase/Vercel stack — before proposing any fix.

Use when review feedback lands on a Maycrest deliverable — before implementing any suggestion, especially an unclear or questionable one — demanding verification over performative agreement.

Use when completing a task or major feature on a Maycrest build, or before merging — dispatch an independent reviewer subagent as the internal QA gate before the client sees the work.

Use when a Maycrest engagement branch is complete and tests pass — gating the handoff: verify the suite first, then choose merge, PR, keep, or discard, and produce a short client handoff package.

Use when a Maycrest client request needs scoping before any build starts — turns a fuzzy ask into an approved written spec through collaborative dialogue, before code or scaffolding.

Use when building any Maycrest client deliverable, feature, or bug fix — before writing implementation code.

Use when about to claim a Maycrest deliverable is complete, fixed, or passing — before committing, opening a PR, or telling the client, requiring fresh verification evidence first.

Use when starting client engagement work that needs isolation — an isolated workspace per engagement so one client's branch never bleeds into another's, via native tools or a git worktree fallback.

Use when coordinating a multi-agent development pipeline from spec to ship — running the Sloth Flow workflow, sequencing dev-QA loops, or enforcing quality gates across agents.

Use when adding AI features with the Anthropic SDK and Claude — chatbots, RAG or semantic search with Supabase pgvector, streaming responses, tool use, or classification and extraction.

Use when running the full app lifecycle on the Expo + Supabase + Stripe stack — scoping a client app, planning a SaaS or multi-tenant platform, or moving from discovery through EAS build.

Use when an engineering or technical task needs routing — backend, frontend, mobile, DevOps, AI, data, or security work — and you want the right specialist picked and deployed.

Use when architecting a Supabase backend — designing schemas, writing Edge Functions or RLS policies, designing APIs, wiring Stripe webhooks, or planning multi-tenant systems and auth flows.

Use when designing or reviewing a Postgres/Supabase schema, writing RLS policies, planning migrations or indexes, choosing a Supabase plan, or diagnosing slow queries.

Use when setting up or deploying backend infrastructure — configuring a Supabase project or Vercel pipeline, wiring Apple sign-in and APNs, managing secrets and environments, or standing up CI/CD.

Use when improving retention or engagement in an Expo/React Native app — designing onboarding flows, habit loops, gamification and reward systems, or push notification sequences to reduce churn.

Use when consolidating data from multiple sources into Supabase — building ETL pipelines, merging or syncing sources, cleaning and validating records, or producing reports for dashboards.

Use when building analytics infrastructure on Supabase Postgres — designing data pipelines and dashboard backends, materialized views, event tracking, or data exports and audit logs.

Use when researching a topic in depth — competitive or market analysis, technology evaluation, feasibility assessment, or any structured deep dive that needs a sourced brief.

Use when working on developer relations — auditing developer experience, writing tutorials or technical content, building community, or feeding developer feedback into the roadmap.

Use when automating deployment pipelines — CI/CD with GitHub Actions, Vercel deploys or EAS Build/Submit/Update, Supabase migrations in CI, or rollback and release workflows.

Use when making sense of multi-source user feedback — app store reviews, support tickets, user interviews, or NPS data — to find patterns and turn them into ranked product priorities.

Use when building frontend UI for Expo (React Native) mobile or Next.js web apps — screens, components, navigation, NativeWind or Tailwind styling, animations — or shipping mobile via EAS Build.

Use when designing or implementing game audio — sound design and SFX pipelines, adaptive music systems, FMOD or Wwise event integration, spatial audio, or audio performance budgets.

Use when designing game systems — gameplay loops, mechanics, economy balance, player progression, or GDDs — or applying gamification to apps like SlothFit challenges and TIE threat simulations.

Use when developing in the Godot 4 engine — writing GDScript 2.0 or Godot C#, designing signal architectures and scene composition, or optimizing Godot performance.

Use when designing game levels and maps — layouts, encounter sequences, pacing arcs, blockout specs, and environmental storytelling — or structuring app onboarding as spatial experiences.

Use when designing the Maycrest website or TIE Platform — building pages, components, or dashboards, choosing fonts and colors, applying the brand aesthetic, or seeking web design inspiration.

Use when building an Expo or React Native mobile app — screens, EAS Build, App Store or Google Play submission, or native features like push notifications, deep linking, or biometrics.

Use when evaluating an LLM or ML model — auditing prompt quality, testing Claude outputs for consistency or safety, running calibration or fairness checks, or producing an audit-grade model report.

Use when writing branching dialogue, designing story or lore systems, defining character voice and arcs, or crafting narrative-driven app onboarding.

Use when governing AI costs and reliability — setting up multi-model LLM routing, circuit breakers and fallbacks, shadow testing with LLM-as-a-Judge, or tracking token spend per execution.

Use when building a prototype, MVP, or proof of concept fast — validating an idea, spinning up a quick demo or landing page, or shipping a testable v0 within days.

Use when writing production-quality Expo or Next.js code — refactoring, implementing a complex feature, writing tests, debugging a hard problem, or improving TypeScript types and performance.

Use when building Metal GPU rendering pipelines — spatial rendering for macOS or Vision Pro, high-performance 3D visualization, compute shaders, stereo streaming, or Metal profiling.

Use when prioritizing a sprint, scoring a feature backlog with ICE or RICE, planning a release, or deciding what to build next across projects.

Use when writing custom shaders, optimizing VFX or particle systems, defining LOD and asset budgets, or profiling GPU performance to bridge art and engine in Unity, Unreal, or Godot.

Use when writing developer documentation — a README, API or Edge Function reference, setup guide, runbook, schema docs, or a technical spec for an app or platform.