AI Agent Skills: Overview & Mental Model | Nexload SDK

Equip Cursor, Claude Code, Antigravity, Windsurf, and Copilot with automated knowledge of Nexload SDK standards.

Nexload Agent Skills are lightweight, standardized instructions adhering to the Agent Skills Specification v1.0. When installed in your repository, your AI assistant gains immediate, zero-shot compliance with production invariants:

• Zero-Hallucination Health Routes: Avoid database cascade failures in liveness probes, configure zero-cache Next.js route handlers, and monitor Linux cgroup resource limits accurately.
• Deterministic Payload Schemas: Craft unified Payload collection fields and Zod schemas without double-declaring types or violating synchronous validation boundaries.
• Clean Architecture & Boundary Narrowing: Write minimal, reviewable TypeScript changes that respect trust boundaries and package export seams.

---

How Nexload Skills Compose (The 3-Tier Mental Model)

Rather than treating skills as an unstructured list of tools, the Nexload ecosystem is engineered as a three-tier symbiotic architecture:

┌─────────────────────────────────────────────────────────────────┐
│               TIER 1: THE COGNITIVE BRAIN                       │
│             nexload-reasoning-* (6+1 Graph)                     │
│  Frames problems, investigates reality, ideates mechanisms,    │
│  synthesizes architecture, evaluates trade-offs, plans delivery │
└───────────────────────────────┬─────────────────────────────────┘
                                │ (Handoffs & Boundaries)
                                ▼
┌─────────────────────────────────────────────────────────────────┐
│              TIER 2: THE ENGINEERING HANDS                      │
│            nexload-code | nexload-package | ...                 │
│  Executes precise TypeScript changes, enforces type safety,     │
│  maintains React render purity, preserves UI, conducts CTO review│
└───────────────────────────────┬─────────────────────────────────┘
                                │ (Domain Contracts)
                                ▼
┌─────────────────────────────────────────────────────────────────┐
│             TIER 3: DOMAIN & PACKAGE SPECIALISTS                │
│             healthcheck-* | payload-*                           │
│  Applies deep, battle-tested operational rules for service      │
│  reliability, Lexical editors, custom fields, and operations   │
└─────────────────────────────────────────────────────────────────┘

1. The Cognitive Brain (nexload-reasoning-*): Decides what problem to solve, how deeply to reason, and which trade-offs to accept through an orchestrated cognitive graph.
2. The Engineering Hands (nexload-*): Knows how to write correct TypeScript, maintain package boundaries, and prevent architectural debt.
3. The Domain Specialists (healthcheck-*, payload-*): Owns package-specific invariants, security models, and lifecycle behaviors.

---

Explore the Agent Ecosystem

Quickstart & Setup Run the One-Shot Adoption Prompt or install skills selectively using the official CLI.

Cognitive Graph (Tier 1) Explore the Kernel and 6 cognitive specialists: Discovery, Investigation, Ideation, Design, Evaluation, Execution.

Engineering Standards (Tier 2) Enforce foundational standards: TypeScript narrowing, React render purity, dual ESM/CJS packaging, and CTO review.

Domain Specialists (Tier 3) Battle-tested skills for @nexload-sdk/healthcheck and @nexload-sdk/payload suite packages.

Governance & Protocols The Inviolable Constitution (8 Laws) and Sparse Pointer Handoff Protocol for cross-agent coordination.

Package Catalog Browse all published runtime packages powering these skills.