From arch-guard
This skill should be used when the user asks "review implementation", "check design compliance", "impl-review", "does this code match the architecture?", "responsibility boundary check", or wants to verify implementation code against architecture docs — checking component responsibility boundaries, interface contract compliance, and reference rule compliance.
How this skill is triggered — by the user, by Claude, or both
Slash command
/arch-guard:impl-reviewThis skill is limited to the following tools:
The summary Claude sees in its skill listing — used to decide when to auto-load this skill
Reviews implementation code against `arch-guard.json` rules and architecture documentation to check design compliance.
Reviews implementation code against arch-guard.json rules and architecture documentation to check design compliance.
/impl-review MyApp.Execution.Workflow
/impl-review src/MyApp.Execution.Workflow/Services/StepExecutor.cs
Without arguments, targets recently changed files (git diff --name-only HEAD~1).
Read arch-guard.json from the project root before starting. This is the single source of truth for all rules.
git diff --name-only HEAD~1 to find recently changed filesconfig.layers[].pattern to identify the layerconfig.docs.architecture is set, read the architecture document to find the relevant section for this layer/componentUsing the architecture documentation (from config.docs.architecture), verify that each component stays within its defined responsibilities:
Using arch-guard.json:
config.references.forbidden[] for violationsconfig.references.cross_layer[] for unauthorized cross-layer references## Impl-Review Report
### Target: {project_or_file}
### Layer: L{level} {layer_name}
### Architecture Reference: {doc_path} {section}
### Compliant
- OK {description of compliant behavior}
### Violations
- VIOLATION {description of violation}
Reference: {config rule or architecture doc section}
### Warnings
- WARNING {description of concern}
Reference: {rule or section}
### Next Steps
(based on results)
Recommend the next arch-guard skill based on results. Always use slash command format:
| Result | Recommendation |
|---|---|
| Violations exist | Fix then → /impl-review {same target} (re-verify) |
| Warnings only | /arch-check — cross-check layer boundary compliance |
| All compliant | /track — check phase progress, or /integration-map {project} — trace change impact |
Provides behavioral guidelines to reduce common LLM coding mistakes, focusing on simplicity, surgical changes, assumption surfacing, and verifiable success criteria.
Searches, retrieves, and installs Agent Skills from prompts.chat registry using MCP tools like search_skills and get_skill. Activates for finding skills, browsing catalogs, or extending Claude.
npx claudepluginhub jaebit/claudemate --plugin arch-guard