root.node Skills

The architecture system for Claude.

27 Skills that diagnose, build, and optimize every architectural layer across chat-Projects, conversations, and Claude Code repos.

---

Quick Start

These Skills install into two surfaces: chat-Projects (Claude.ai web app or Claude desktop app) and Claude Code (the execution surface). Most Skills are chat-Project-only. Three are Claude Code-only. Two operate in both.

For chat-Projects. Go to the latest release — the catalog umbrella, with every Skill indexed by category. Download the individual -cp.zip files for the Skills you want (no unzipping required), or grab the all-Skills bundle rootnode-catalog-v4.0.zip and unzip it once. Upload the -cp.zip files in Settings → Capabilities → Skills. Skills install once and become available across every Project in your Claude account.

For Claude Code. Go to the latest release, download the -cc.zip files for the Skills you want (or the all-Skills bundle rootnode-catalog-v4.0.zip, unzipped once), then extract the -cc.zip archives into ~/.claude/skills/ (user-level — available across every repo) or .claude/skills/ (per-repo). Each -cc.zip expands to a folder matching the Skill name. Skills auto-activate from natural language prompts inside Claude Code sessions.

The two dual-surface Skills (rootnode-skill-builder and rootnode-cc-design) publish both -cp.zip and -cc.zip variants — download the one matching your install target.

[!NOTE] Activate any Skill with natural language — no commands, no special syntax. Install all 27 for the full architecture system; every Skill also works standalone.

[!TIP] Calibrated for Claude Opus 5 + Sonnet 5 (dual-primary), v4.0. Fable 5 integrated-aware; Opus 4.8 fallback-graceful (Skills produce correct-shape output there when classifier-flagged requests fall back silently); Sonnet 4.6 legacy-graceful; Haiku 4.5 with extended thinking enabled. Most Skills work fully across the current model landscape; a few produce best results on Opus 5 or Sonnet 5 at high effort. See Model Compatibility for details.

---

Why Architecture Matters

The Claude ecosystem has a layered architecture. How content is distributed across those layers determines whether Claude performs at its potential or works against itself. The hard part isn't building a Project or wiring a Claude Code repo — it's seeing what's quietly working against you.

Architectural inefficiencies don't announce themselves. They surface as vague symptoms: output that drifts, instructions that get ignored intermittently, quality that degrades over long conversations, autonomous agents that escalate work they shouldn't or barrel through halts they should respect. These Skills surface those structural issues, triage them by impact, and produce targeted fixes.

The trap is that these symptoms look like behavior problems. They're not. A rule the user has rewritten five different ways still gets ignored because the layer it sits in is wrong, not because the rule is unclear. Cross-layer conflicts cause unpredictable output that no amount of prompt rewording will fix. Misplaced enforcement in a Claude Code environment becomes preference instead of guarantee. Structural problems require structural solutions.

These Skills treat the entire ecosystem as a unified system — scoring each layer, detecting cross-layer conflicts, diagnosing behavioral tendencies, gating handoffs to autonomous execution, and building Projects and Claude Code environments that are structurally sound from the start. The system is calibrated to how Claude actually processes context: the behavioral tendencies documented through extensive testing, the loading behavior of each layer, and the architectural patterns that produce reliable output across both surfaces.

---

The Architectural Layers

Two surfaces, layered architecture across each. Skills are mapped to the layers they impact. A Skill that builds knowledge files appears on the Knowledge Files layer because that's the layer it modifies; a Skill that audits five layers at once appears on each of them.

One Skill spans every layer. rootnode-full-stack-audit runs Project audit + Global audit + Cross-Layer Alignment in a single pass — the only Skill in the catalog that operates across scopes rather than within one. See Cross-Layer in Core Skills for details.

Global Scope

Five layers shared across every chat-Project: User Preferences, Styles, Global Memory, Skills, and MCP Connectors. These define the foundation Claude works with before any specific Project loads. Edits here cascade across your entire Claude environment, which is why structural problems at the Global level are some of the most consequential — and most invisible — issues a user can have.

Project Scope

Three layers that shape every conversation in a chat-Project: Project Memory (orientation state), Custom Instructions (always-loaded behavioral architecture), and Knowledge Files (searchable reference depth). Most chat-side quality issues originate here — wrong content in wrong layer, instructions that fight each other, retrieval mode pressure from oversized Knowledge Files. The majority of root.node's Core Skills target this scope because it's where structure most directly determines output.

Conversation Scope

Where you interact with the chat-Project. Most Skills activate here from natural language — block selection, domain packs, identity templates, reasoning approaches, output formats. Two layers of continuity sit on top of the conversation:

Bridges connect within the chat-side ecosystem. project-brief captures Project context for use in other Projects; session-handoff preserves work state across separate conversations. Both produce artifacts