Bossku-AI

Use this for Hyperframes video composition work, including HTML-based scene authoring, preview/render workflows, CLI commands, agent-friendly composition updates, and GSAP-assisted animation inside Hyperframes projects.

Use when a Codex workspace needs BosskuAI routing, repo-local memory, skill selection, or quality gates without loading the full toolkit.

Use this for cloud/server/API/model cost control, token budgets, queue sizing, storage costs, vendor spend, and unit economics.

Use this for support SOPs, onboarding, ticket triage, customer success playbooks, retention, churn reduction, and feedback loops.

Use this for agent eval design, routing tests, retrieval tests, LLM quality cases, regression harnesses, scorecards, and continuous agent improvement.

Use this for Malaysia PDPA-aware privacy review, data minimization, consent, retention, user rights, vendor processors, and privacy-safe SaaS operations.

Use this for logs, metrics, tracing, alerts, SLOs, health checks, dashboards, incident detection, and production reliability instrumentation.

Use this for prompt injection, tool abuse, memory poisoning, untrusted document handling, agent permissions, and AI workspace security.

Use this for automated testing strategy, Playwright/Cypress, feature tests, integration tests, regression suites, fixtures, CI gates, and release confidence.

Use this for SaaS subscription lifecycle, invoices, failed payments, dunning, tax/receipt flows, plan changes, entitlement gating, and billing operations.

Use this for multi-tenant data isolation, organization scoping, cross-tenant leaks, authorization boundaries, row-level access, and SaaS tenant security review.

Use this when the user wants BosskuAI to behave like an expert AI cofounder across product, engineering, UI/UX, security, SEO/GEO, GTM, marketing, sales, prioritization, and next-step decision making.

Use this when starting any non-trivial coding work to apply the four engineering principles — Think Before Coding, Simplicity First, Surgical Changes, Goal-Driven Execution. Pairs with bosskuai-coding-best-practices (which covers code-quality detail) and bosskuai-rigorous-code-review (which audits the result).

Use this for product strategy, feature shaping, PRD review, scope refinement, prioritization, user-value analysis, and go-to-market implications of product decisions.

Use this for creating immersive 3D websites, WebGL experiences, Three.js/React Three Fiber scenes, scroll-driven 3D animations, GSAP-powered 3D motion, post-processing effects, Spline integrations, interactive particle systems, and Awwwards-quality 3D web experiences.

Use this for securing AI-agent workspaces themselves, including instruction files, MCP setups, memory, external content, prompt-injection surfaces, and least-privilege configuration.

Use this for recommending which AI model fits a specific task based on reasoning depth, speed, tool use, coding needs, multimodality, cost sensitivity, and reliability tradeoffs.

Use this for instrumentation strategy, event design, funnels, experimentation, north-star metrics, KPI definitions, attribution logic, and making sure product decisions can be measured reliably.

Use this for API contract design across REST, GraphQL, and event-driven interfaces, including resource modeling, versioning, errors, pagination, idempotency, and integration-facing correctness.

Use this skill for browser-level testing, UI verification, automated user flow testing, visual regression, and scraping JavaScript-rendered pages. Covers QA smoke tests through full user journey automation.

Use this for bug hunts, regression analysis, suspicious changes, failure-path review, finding likely defects before shipping, and deep incident investigation using logs, DB state, queues, and runtime evidence.

Use this for workflow review, state transitions, approval and assignment design, edge cases, rule consistency, and finding hidden business logic flaws in products or systems.

Deprecated compatibility alias for bosskuai-token-saver. Use only when the user explicitly says caveman mode or references the old caveman skill.

Use this to analyze a repository and recommend Claude Code setup improvements, including CLAUDE.md, MCP servers, hooks, skills, commands, permissions, and onboarding configuration.

Use this to audit, maintain, and improve CLAUDE.md and Claude-specific rule files, including instruction quality, session learning capture, overlap control, and cross-tool consistency with BosskuAI memory.

Use this for safe code modernization, structural cleanup, legacy refactors, and revamps that should still respect the current codebase structure and avoid unnecessary churn.

Use this for tracing execution paths, mapping call chains and module boundaries, identifying side effects, and summarizing how code actually behaves at runtime. Use bosskuai-project-understanding first when the project purpose and stack are still unknown.

Use this skill for structured competitor tracking — feature comparison matrices, pricing intelligence, messaging changes, hiring signals, and strategic implications. Repeatable, cadence-friendly monitoring with delta detection.

Use this for content calendar planning, platform-specific posts, hooks, CTAs, content pillars, Malaysian mixed Malay-English tone, repurposing, and publishing cadence.

Use this when a task risks hitting model context, token limits, or tight usage/quota mid-process. Stop cleanly, write progress into ai-assistant/memory/active-continuation.md, pair with bosskuai-ai-model-selection to recommend which model should complete the remaining work, and tell the user explicitly to open a fresh chat/session with that model. Also contains token-budget guidance for keeping agent sessions efficient.

Use this when the current model is stuck, low-confidence, missing a capability, or repeating failed attempts. It defines how to bring in another model, tool surface, or session for scoped assistance across Claude, Codex, and Cursor without losing ownership of the task.

Use this skill for user interview planning, customer discovery, survey design, interview transcript analysis, and persona building from primary research. Gathers and structures the evidence that product strategy acts on.

Use this for cybersecurity review, privacy, abuse-case analysis, auth and authorization concerns, trust boundaries, fraud risk, and operational risk evaluation.

Use this for data modeling, schema boundaries, migrations, warehouses, analytics pipelines, ownership of entities, retention, and correctness tradeoffs in data-intensive systems.

Use this for SQL and NoSQL database design across MariaDB, MySQL, PostgreSQL, SQLite, MongoDB, indexing, transactions, migrations, constraints, query plans, and data correctness.

Use this skill for in-depth investigation tasks — multi-source research, due diligence, technology evaluation, and evidence synthesis with citations. Produces sourced executive summaries, not strategy.

Use this for creating, auditing, or applying design systems — including DESIGN.md generation, design token definition, component specification, brand reference calibration, and systematic design language enforcement.

Use this for CI/CD, containers, deployment workflows, infrastructure as code, secrets handling, environment promotion, rollback design, and delivery reliability across operational systems.

Use this for Dockerfile and Docker Compose setup or review, including one-command `docker compose up -d` startup, `.env` configuration, no hardcoded credentials in YAML, bind-mounted local volume sync, explicit networks, service health checks, non-root images, and containerized dev/prod workflows.

Use this for disciplined software implementation work that should follow planning-first execution, test-guided development, review-before-finalization, and explicit verification.