Spec Driven Development
MCP Server that facilitates spec-driven development workflows, not just Vibe Coding.
Description from the maker (original English)
Render SwiftUI #Preview as PNG images, so an AI agent sees the views it changes
These three are never merged into one score: opinions, measurements and popularity are different things.
We have not written a plain-language summary for this one yet. Below are the facts we read from the registry.
Pick the AI app you use, then copy the settings.
Which app do you use?
Pick your app above. We remember it, so next time the button says “Install in …”.
Download the maker’s .mcpb file
The link goes to the maker’s own file; we do not host or repackage it. Open the file and Claude Desktop installs it.
Run this in your terminal
Your browser asks to open the app, then the app shows the settings and asks you to confirm. Key values are left empty for you to fill in.
Your browser asks to open the app, then the app shows the settings and asks you to confirm. Key values are left empty for you to fill in.
This one runs on your computer, and ChatGPT can only connect to servers on the internet. Pick another app.
Your choice stays on this device in a cookie; no account needed.
These settings are generated from the maker’s registry entry. If the README says otherwise, follow the README.
Runs on your computer; the tool list appears in your AI app after you install it.
We check information anyone can see; we have not read the code. Checks that do not apply are left out of the count. · Checked Nov 6, 2026
The registry name was verified with the GitHub account BarredEwe, which owns the repo
We have not read this repo from GitHub yet
We have not read this repo from GitHub yet
This one runs on your computer; there is no internet endpoint
Runs on your computer and uses local settings instead
Runs on your computer; we can only read tool lists from internet endpoints
Compared with every active server in this category in the registry. Stars count only when the publisher owns the repo.
It passed 1 of 1 applicable basic checks. We have not read the code. Before adding a key or token, see what access the server asks for and use a key with only the permissions you need.
The registry does not list a required key, though some servers ask you to sign in on first use.
It runs on your computer. Your AI app starts it when needed, and data does not pass through the maker’s servers (unless it calls an online service itself).
The GitHub repo last changed –, and the latest version (6.6.0) reached the registry 7 days ago.
Apps that support MCP, such as Claude Desktop, Claude Code, Cursor and VS Code. You also need MCPB on your computer.
In the same category: Spec Driven Development, Remote Desktop Commander, dotnet-mcp
Servers in the same category from other makers, sorted by stars.
Reading a card:★ GitHub starsiThree pillars: reviews · checks · usageiUpdated = latest version published to the registryi
MCP Server that facilitates spec-driven development workflows, not just Vibe Coding.
Remote Desktop Commander: use files, terminal and processes on your own computers from any AI app.
MCP server for .NET symbols, XAML, source generators, and restricted workspace edit
Copied from the maker’s repo as written. PopNx MCP has not reviewed or endorsed it. HTML, images and scripts are removed. Links lead off this site.
Quick Start · Installation · Usage · AI agents · Configuration · Docs · Website
Your previews are already tests. Prefire turns the #Preview blocks you write anyway into snapshot tests, a Playbook catalog and images your AI agent can see. You don't write a line of test code.
📸 Snapshot tests Every preview becomes a snapshot test on the devices you pick. A changed padding fails the build, not the release.
How it works →
📚 Playbook A catalog of every component and user story inside your app, with their states and render times.
See Playbook →
🤖 Renders for AI agents Claude Code, Cursor and Codex render your previews through an MCP server and see the UI they just changed.
Set up →
🤖 Using an AI agent?
brew install barredewe/prefire/prefire, thenclaude mcp add prefire -- prefire mcpin your project. The agent renders SwiftUI previews on any iPhone or iPad, light or dark, without Xcode Previews. Setup →
#Preview, @Previewableprefire render and an MCP serverUIView/UIViewController and NSView/NSViewController📦 Example project available at: Prefire Example
// Package.swift
dependencies: [ .package(url: "https://github.com/BarredEwe/Prefire.git", from: "6.3.0"), .package(url: "https://github.com/pointfreeco/swift-snapshot-testing", from: "1.18.0"), ], .testTarget( dependencies: [ .product(name: "Prefire", package: "Prefire"), .product(name: "SnapshotTesting", package: "swift-snapshot-testing"), ], plugins: [ .plugin(name: "PrefireTestsPlugin", package: "Prefire") ] )
#Preview#Preview {
Button("Submit") }
Just run the test target 🚀 — Prefire will auto-generate snapshots based on your previews.
💡 If your test target is empty, Prefire will still generate files and snapshot code during build.
| Method | How |
|---|---|
| SPM Plugin | PrefireTestsPlugin / PrefirePlaybookPlugin in Package.swift |
| Xcode Build Tool Plugin | Add PrefireTestsPlugin to the test target's build phases |
| CLI | brew install barredewe/prefire/prefire |
| Mint (*Beta*) | mint install BarredEwe/Prefire |
| GitHub Actions / CI | Any of the above, see Troubleshooting |
Step-by-step setup is in the Installation guide.
Prefire finds every `#Preview` automatically. For PreviewProvider, add the PrefireProvider protocol:
struct Text_Previews: PreviewProvider, PrefireProvider {
static var previews: some View { ... } }
Exclude a preview with .prefireIgnored():
#Preview {
Text("") .prefireIgnored() }
To opt in instead of opting out, set preview_default_enabled: false in .prefire.yml and mark the previews you want with .prefireEnabled():
#Preview {
Text("") .prefireEnabled() }
PrefireTestsPlugin generates the tests into the DerivedData folder when the test target builds. Run the test target as usual.
Tune the timing and precision of a single snapshot:
#Preview {
TestView() .snapshot(delay: 0.3, precision: 0.95, perceptualPrecision: 0.98) }
For previews that load or animate, wait instead of guessing a delay:
.snapshot(waitForIdle: true) // until nothing changes on screen
.snapshot(waitUntil: { viewModel.isLoaded }) // until a condition is true
How waiting works, its limits and how the modifiers combine: Snapshot timing and precision.
Snapshots are compared with swift-snapshot-testing.
Show all previews as an interactive catalog with PlaybookView:
isComponent: true lists all the ViewsisComponent: false groups them by UserStoryimport Prefire
struct ContentView: View { var body: some View { PlaybookView(isComponent: true, previewModels: PreviewModels.models) } }
To keep previews and PreviewModels out of release builds, see Excluding previews from release builds.
Connect previews into one Flow with .previewUserStory. Playbook shows them together, in order:
static var previews: some View {
PrefireView() .previewUserStory(.auth) }
static var previews: some View { AuthView() .previewUserStory(.auth) }
If a preview contains more than one View, mark the State of each:
static var previews: some View {
TestView("Default")
TestView("Loading") .previewState(.loading) }
Prefire expands each argument of a native SwiftUI parameterized preview into its own snapshot and Playbook preview:
#Preview("TextView", traits: .sizeThatFitsLayout, arguments: ["- A", "- B", "- C"]) { suffix in
Text("1 \(suffix)") }
| Feature | Modifier |
|---|---|
| Include in snapshot | .prefireEnabled() |
| Exclude from snapshot | .prefireIgnored() |
| Group in a flow | .previewUserStory(.auth) |
| Mark a UI state | .previewState(.error) |
| Customize snapshot | .snapshot(delay: 0.3, precision: 0.95) |
| Wait until the preview settles | .snapshot(waitForIdle: true, timeout: 2.0) |
| Wait for a condition | .snapshot(waitUntil: { viewModel.isLoaded }) |
| Parameterized preview | #Preview(..., arguments: values) |
# Generate snapshot tests
prefire tests
prefire playbook
Run prefire tests --help or prefire playbook --help for more options.
Prefire renders previews in its snapshot tests, built anew and on the devices you choose, so it can show them to an AI agent as images, which Xcode Previews can't do reliably outside Xcode.
# Render previews to PNG + JSON
prefire render "Sign in" --device "iPhone 16 Pro" --color-scheme dark
prefire board
Add the MCP server to your agent, from the folder with .prefire.yml:
| Agent | Install |
|---|---|
| Claude Code | claude mcp add prefire -- prefire mcp, or the plugin with a skill: /plugin marketplace add BarredEwe/Prefire then /plugin install prefire@prefire |
| Cursor | |
| Codex | codex mcp add prefire -- prefire mcp |
| VS Code | code --add-mcp '{"name":"prefire","command":"prefire","args":["mcp"]}' |
An agent then calls list_previews, render_previews and build_board. Set the scheme to build in render_configuration of .prefire.yml; see the AI guide. Why not Xcode's own RenderPreview? Prefire needs no running Xcode, renders many previews in one build, on the devices, color scheme and Dynamic Type you ask for, and works in CI.
.prefire.ymltest_configuration:
target: MyApp
playbook_configuration: preview_default_enabled: true
All options are in the Configuration guide. To change the generated code, see Templates.
#Preview and PreviewProvider blocks and the .prefireEnabled() / .prefireIgnored() modifiersStencil templates, respecting .prefire.ymlUserStory and State into PreviewModels.generated.swiftOnly the beginning is shown; read the rest in the repo