architecture-compass
Architectural thinking partner for an existing repository — scans the codebase, conducts a structured interview, agrees on current architectural state and…
Enforce architectural rules when generating or modifying code, and validate proposed designs before approval (design mode). Defaults to clean architecture; supports any architecture style via the architecture-refiner. Validates layer responsibilities, dependency direction, and
$ npx -y skills add techygarg/lattice --skill architecture --agent claude-codeHow it fires
How this skill gets triggered: by you, by Claude, or both.
/architectureContext preview
The summary Claude sees to decide when to auto-load this skill.
Enforce architectural rules when generating or modifying code, and validate proposed designs before approval (design mode). Defaults to clean architecture; supports any architecture style via the architecture-refiner. Validates layer responsibilities, dependency direction, and
name: architecture description: "Enforce architectural rules when generating or modifying code, and validate proposed designs before approval (design mode). Defaults to clean architecture; supports any architecture style via the architecture-refiner. Validates layer responsibilities, dependency direction, and structural constraints using the loaded architecture rules. Use when generating code, reviewing architecture, creating new files, or when the user mentions 'architecture', 'layers', 'structure', 'dependency rules', 'hexagonal architecture', 'ports and adapters', 'modular monolith', or 'onion architecture'. Also use when reviewing generated code for structural compliance."
**Step 1 — Determine mode:**
1. Read `.lattice/config.yaml` in repo root 2. Check `architecture_mode` key
**Step 2 — Load enforce rule:**
**Step 3 — Load architecture content:**
1. Check `paths.architecture` in `.lattice/config.yaml` for a custom document path. 2. If a document exists at that path → read it and check its YAML frontmatter for `mode`:
3. If the path is configured but no document exists at it → tell the user which configured path is missing, then read `./references/clean-architecture-defaults.md`. 4. If there is no configured path → read `./references/clean-architecture-defaults.md`.
1. Check `paths.architecture` in `.lattice/config.yaml` for team architecture doc 2. If found → read it. Sole reference — no default. 3. If not found → surface: "No architecture document found. Run `/architecture-refiner` and select your architecture style to define your team's standards."
**Step 4 — Language adaptation:**
If `paths.language_idioms` is set in the config and the document exists, read its **"Dependency Management"** section and adapt dependency direction enforcement to language idioms (e.g., Go interface-at-consumer, Java DI containers, Rust trait bounds). Language idioms take precedence over pseudocode defaults.
**STOP after generating each component.** Read the **Self-Validation Checklist** and **Anti-Pattern Scan** from the loaded enforce rule (clean-architecture.md or custom-architecture.md) and apply them.
**Project-specific checks:** If architecture content doc (loaded in Step 3) contains a **Validation Checklist** section (§6), apply those checks as additional project-specific validation after the enforce rule checklist.
When invoked during design — no code is being written; a planning molecule is validating a proposed design artifact — apply the same enforce rule as a forward-looking check:
1. Take the proposed artifact (component list, layer assignment, data flow, or contract set) as the unit of validation. 2. Evaluate it against the **Self-Validation Checklist** and **Anti-Pattern Scan** from the loaded enforce rule, plus the project-specific Validation Checklist if present — before the artifact is presented for user approval. 3. Report violations as concrete findings on the artifact ("Component X reaches from layer A to layer C, skipping B"), not generic advice. Resolve them through the design. 4. **STOP:** do not skip checklist evaluation on the grounds that no code exists yet — the proposed structure is what gets validated.
Composable AI skills that teach assistants structured thinking — design-first, context-aware, and architecture-guided.
Repo: techygarg/lattice
Architectural thinking partner for an existing repository — scans the codebase, conducts a structured interview, agrees on current architectural state and…
Facilitate a structured conversation to define architecture principles for a repository. Supports multiple architecture styles: clean architecture (default),…
Investigate, reproduce, and safely fix a bug with regression protection. Composes context, diagnosis, architecture, code quality, and testing guardrails into a…
Facilitate a structured conversation to define clean code principles for a repository. Produces a formal clean-code.md document that the clean-code atom will…
Apply clean code principles when generating or modifying implementation code. Enforces function focus, naming clarity, complexity management, error handling,…
Generate implementation code from an approved design blueprint or verbal requirements. Composes context anchoring, architecture, clean code, DDD, security, and…