refactor(12/12): add snapshot test fixtures and benchmarks

[cameroncooke]

Summary

This is PR 12 of 12, the final PR in the stacked series that decouples the rendering pipeline from MCP transport. Depends on PR 11.

Adds the snapshot test suite and performance benchmarks that validate the entire rendering pipeline end-to-end. These are large in line count but are almost entirely test fixtures (expected output files) and benchmark scripts.

Snapshot test suite (159 files, ~5700 lines)

The snapshot test infrastructure captures the rendered output of tool invocations and compares against expected fixtures. This provides regression protection for the rendering pipeline -- any change to event formatting, diagnostic grouping, or output ordering will be caught by a fixture mismatch.

Test harness (src/snapshot-tests/):

• harness.ts: Core test runner that invokes tools with mock executors and captures rendered output
• fixture-io.ts: Reads/writes fixture files, handles normalization (timestamps, paths, UUIDs)
• flowdeck-fixture-io.ts: Flowdeck-specific fixture handling
• normalize.ts: Output normalization for stable comparisons across environments
• resource-harness.ts: Resource-specific snapshot testing

Fixtures: Expected output files for each tool covering success, error, and edge case scenarios. These serve as living documentation of what each tool's output looks like.

Benchmarks (~3300 lines)

Performance benchmarks for the rendering pipeline and xcodebuild parsing:

• Parser throughput: lines/second for xcodebuild output parsing
• Render session performance: events/second for text and JSON strategies
• End-to-end tool invocation timing

These benchmarks establish baselines and can be run in CI to catch performance regressions.

Note for reviewers

This PR is large by line count but low in conceptual complexity. The fixture files are auto-generated expected outputs. The benchmark scripts are straightforward timing loops. The meaningful code is the ~500 lines of test harness infrastructure.

Stack navigation

• PR 1-11/12: All code and configuration changes
• PR 12/12 (this PR): Snapshot tests and benchmarks

Test plan

• npx vitest run passes -- snapshot tests match expected fixtures
• npx vitest run --config vitest.snapshot.config.ts runs snapshot suite specifically
• Benchmarks execute without errors (performance numbers are informational)

[cursor]

Cursor Bugbot has reviewed your changes and found 2 potential issues.

Autofix Details

Bugbot Autofix prepared fixes for both issues found in the latest run.

• ✅ Fixed: Fixture has un-normalized username causing cross-environment test failures
  • Added normalization for USER, ALTERNATE_OWNER, INSTALL_OWNER, and VERSION_INFO_BUILDER build settings to replace usernames with , and updated the fixture file accordingly.
• ✅ Fixed: Developer team ID committed in fixture file
  • Added normalization for DEVELOPMENT_TEAM build setting to replace team ID with , and updated the fixture file to remove the hardcoded team ID.

Or push these changes by commenting:

@cursor push d89a7eafb2

Preview (d89a7eafb2)

diff --git a/src/snapshot-tests/__fixtures__/project-discovery/show-build-settings--success.txt b/src/snapshot-tests/__fixtures__/project-discovery/show-build-settings--success.txt
--- a/src/snapshot-tests/__fixtures__/project-discovery/show-build-settings--success.txt
+++ b/src/snapshot-tests/__fixtures__/project-discovery/show-build-settings--success.txt
@@ -15,7 +15,7 @@
       ALLOW_TARGET_PLATFORM_SPECIALIZATION = NO
       ALTERNATE_GROUP = staff
       ALTERNATE_MODE = u+w,go-w,a+rX
-      ALTERNATE_OWNER = cameroncooke
+      ALTERNATE_OWNER = <redacted>
       ALTERNATIVE_DISTRIBUTION_WEB = NO
       ALWAYS_EMBED_SWIFT_STANDARD_LIBRARIES = NO
       ALWAYS_SEARCH_USER_PATHS = NO
@@ -156,7 +156,7 @@
       DEVELOPER_TOOLS_DIR = /Applications/Xcode-26.4.0.app/Contents/Developer/Tools
       DEVELOPER_USR_DIR = /Applications/Xcode-26.4.0.app/Contents/Developer/usr
       DEVELOPMENT_LANGUAGE = en
-      DEVELOPMENT_TEAM = BR6WD3M6ZD
+      DEVELOPMENT_TEAM = <redacted>
       DIAGNOSE_MISSING_TARGET_DEPENDENCIES = YES
       DIFF = /usr/bin/diff
       DOCUMENTATION_FOLDER_PATH = CalculatorApp.app/en.lproj/Documentation
@@ -293,7 +293,7 @@
       INSTALL_DIR = /tmp/CalculatorApp.dst/Applications
       INSTALL_GROUP = staff
       INSTALL_MODE_FLAG = u+w,go-w,a+rX
-      INSTALL_OWNER = cameroncooke
+      INSTALL_OWNER = <redacted>
       INSTALL_PATH = /Applications
       INSTALL_ROOT = /tmp/CalculatorApp.dst
       IPHONEOS_DEPLOYMENT_TARGET = 17.0
@@ -568,7 +568,7 @@
       UNLOCALIZED_RESOURCES_FOLDER_PATH_SHALLOW_BUNDLE_NO = CalculatorApp.app/Resources
       UNLOCALIZED_RESOURCES_FOLDER_PATH_SHALLOW_BUNDLE_YES = CalculatorApp.app
       UNSTRIPPED_PRODUCT = NO
-      USER = cameroncooke
+      USER = <redacted>
       USER_APPS_DIR = <HOME>/Applications
       USER_LIBRARY_DIR = <HOME>/Library
       USE_DYNAMIC_NO_PIC = YES
@@ -579,7 +579,7 @@
       VALID_ARCHS = arm64 arm64e armv7 armv7s
       VERBOSE_PBXCP = NO
       VERSIONPLIST_PATH = CalculatorApp.app/version.plist
-      VERSION_INFO_BUILDER = cameroncooke
+      VERSION_INFO_BUILDER = <redacted>
       VERSION_INFO_FILE = CalculatorApp_vers.c
       VERSION_INFO_STRING = "@(#)PROGRAM:CalculatorApp  PROJECT:CalculatorApp-1"
       WATCHOS_DEPLOYMENT_TARGET = 26.4

diff --git a/src/snapshot-tests/normalize.ts b/src/snapshot-tests/normalize.ts
--- a/src/snapshot-tests/normalize.ts
+++ b/src/snapshot-tests/normalize.ts
@@ -47,6 +47,9 @@
 const ACQUIRED_USAGE_ASSERTION_TIME_REGEX =
   /(^\s*)\d{2}:\d{2}:\d{2}( {2}Acquired usage assertion\.)$/gm;
 const BUILD_SETTINGS_PATH_REGEX = /^( {6}PATH = ).+$/gm;
+const BUILD_SETTINGS_USER_REGEX =
+  /^( {6}(?:USER|ALTERNATE_OWNER|INSTALL_OWNER|VERSION_INFO_BUILDER) = ).+$/gm;
+const BUILD_SETTINGS_DEVELOPMENT_TEAM_REGEX = /^( {6}DEVELOPMENT_TEAM = ).+$/gm;
 const TRAILING_WHITESPACE_REGEX = /[ \t]+$/gm;
 
 function sortLinesInBlock(text: string, marker: RegExp): string {
@@ -130,6 +133,8 @@
 
   normalized = normalized.replace(TARGET_DEVICE_IDENTIFIER_REGEX, '$1<UUID>');
   normalized = normalized.replace(BUILD_SETTINGS_PATH_REGEX, '$1<PATH>');
+  normalized = normalized.replace(BUILD_SETTINGS_USER_REGEX, '$1<redacted>');
+  normalized = normalized.replace(BUILD_SETTINGS_DEVELOPMENT_TEAM_REGEX, '$1<redacted>');
   normalized = normalized.replace(CODEX_ARG0_PATH_REGEX, '<HOME>/.codex/tmp/arg0/codex-arg0<ARG0>');
   normalized = normalized.replace(ACQUIRED_USAGE_ASSERTION_TIME_REGEX, '$1<TIME>$2');
   normalized = normalized.replace(

_{This Bugbot Autofix run was free. To enable autofix for future PRs, go to the Cursor dashboard.}

[cursor]

Cursor Bugbot has reviewed your changes and found 2 potential issues.

^{❌ Bugbot Autofix is OFF. To automatically fix reported issues with cloud agents, enable autofix in the Cursor dashboard.}

^{Reviewed by Cursor Bugbot for commit daa650e. Configure here.}

[cameroncooke]

Warning

This pull request is not mergeable via GitHub because a downstack PR is open. Once all requirements are satisfied, merge this PR as a stack on Graphite.
Learn more

• refactor(12/12): add snapshot test fixtures and benchmarks #330 👈 (View in Graphite)
• refactor(11/12): update manifests, config, docs, and example projects #329
• refactor(10/12): rewire CLI, daemon, and MCP server boundaries #328
• refactor(9/12): migrate debugging, swift-package, and remaining tools #327
• refactor(8/12): migrate UI automation tools to event-based handlers #326
• refactor(7/12): migrate device and macOS tools to event-based handlers #325
• refactor(6/12): migrate simulator tools to event-based handlers #324
• refactor(5/12): change handler contract to event-based emission #323
• refactor(4/12): extract build/test utilities, platform steps, and xcodebuild pipeline #322
• refactor(3/12): add rendering engine and output formatting #321
• refactor(2/12): add pipeline event types, parsers, and event builders #320
• refactor(1/12): remove deprecated logging tools #319
• main

This stack of pull requests is managed by Graphite. Learn more about stacking.