Skip to main content

When to Use Anvil

Anvil is powerful but focused. This page helps you understand when it's the right tool.

Anvil is for you if...​

You ship with AI assistance​

Anvil is designed for teams using AI coding tools (Copilot, Claude, Cursor, etc.). If you're not using AI assistance, the ROI is lower—though the architecture safety features still provide value.

You care about architecture​

Anvil shines when you have intentional architectural boundaries:

  • Layer separation (UI → Services → Data)
  • Domain boundaries (Payments shouldn't import User internals)
  • Package boundaries (Public API vs internal implementation)

You want save-time feedback​

Anvil's watch mode gives you instant feedback. If you only want CI-time validation, you can use Anvil in CI mode—but you lose the developer experience benefit.

You value audit trails​

Every Anvil run produces evidence. If compliance, traceability, or post-incident analysis matters to you, this is valuable.

Anvil is NOT for you if...​

You want a linter replacement​

Anvil complements ESLint, not replaces it. Anvil catches structural issues; ESLint catches stylistic and semantic issues.

Use both.

You want test coverage enforcement​

Anvil can check coverage thresholds, but it's not a test runner or coverage tool. Use your existing coverage tooling and integrate with Anvil.

You have no architecture to protect​

If your codebase is a single module with no intentional boundaries, Anvil's architecture checks won't find violations. The anti-pattern detection still helps, but you're not using Anvil's primary value.

You need real-time collaboration​

Anvil is a development tool, not a collaboration platform. It doesn't sync state between developers or provide real-time editing features.

Team Size Considerations​

Solo developers​

Anvil works well for solo devs who:

  • Use AI assistance heavily
  • Want to maintain quality without manual review overhead
  • Value the peace of mind from automated checks

Small teams (2-10)​

Sweet spot. Anvil prevents the "AI generated this, nobody really reviewed it" problem. Save-time validation means issues don't pile up for PR review.

Large teams (10+)​

Anvil scales via CI integration. The architecture safety features become more valuable as more developers (and more AI agents) touch the codebase.

Integration with Existing Tools​

Anvil integrates with your existing workflow, not replaces it:

ToolAnvil's Role
ESLintAnvil runs ESLint as a gate check
PrettierAnvil doesn't touch formatting
Jest/VitestAnvil can gate on test pass/coverage
GitHub ActionsAnvil provides a GitHub Action
VS CodeAnvil provides an extension

Decision Framework​

Ask yourself:

  1. Do I use AI coding tools? → If no, consider skipping Anvil
  2. Do I have architectural boundaries? → If no, value is limited
  3. Do I want save-time feedback? → If no, use CI-only mode
  4. Do I need audit trails? → Anvil provides this automatically

If you answered "yes" to 2+ questions, Anvil will likely add value.

Ready to try it? Quickstart →