retestkit 1.4.1 → 1.5.0

package/openspec/changes/archive/2025-12-18-use-markdown-artifacts/specs/webtest-resources/spec.md

@@ -1,109 +0,0 @@
-# webtest-resources Spec Delta
-
-## MODIFIED Requirements
-
-### Requirement: Resource URI Scheme
-
-The system SHALL expose all webtest artifacts using a `webtest://` URI scheme with hierarchical paths, using markdown format for all human-readable artifacts.
-
-#### Scenario: Analysis root resource is accessible
-
-- **GIVEN** an analysis has been started with analysisId "abc123"
-- **WHEN** client requests resource `webtest://abc123/`
-- **THEN** it SHALL return the analysis `index.md` metadata as markdown with YAML frontmatter
-
-#### Scenario: Crawl index resource is accessible
-
-- **GIVEN** a crawl has completed with crawlId "crawl-001"
-- **WHEN** client requests resource `webtest://abc123/crawls/crawl-001/index.md`
-- **THEN** it SHALL return the crawl index as markdown with YAML frontmatter containing page list and metadata
-
-#### Scenario: Page artifacts are accessible by type
-
-- **GIVEN** a page was captured with pageId "page-001"
-- **WHEN** client requests `webtest://abc123/crawls/crawl-001/pages/page-001/screenshot.png`
-- **THEN** it SHALL return the screenshot image
-- **AND** `snapshot.md` returns accessibility tree as formatted markdown with YAML frontmatter
-- **AND** `dom.html` returns HTML content
-
-#### Scenario: Crawl checkpoint is accessible
-
-- **GIVEN** a crawl is in progress with checkpoint saved
-- **WHEN** client requests `webtest://abc123/crawls/crawl-001/checkpoint.md`
-- **THEN** it SHALL return the checkpoint as markdown with YAML frontmatter containing crawl state
-
-#### Scenario: Analysis report is accessible
-
-- **GIVEN** analyze_app has completed
-- **WHEN** client requests `webtest://abc123/analysis/app-analysis.md`
-- **THEN** it SHALL return the markdown analysis report
-
-#### Scenario: Flows are accessible
-
-- **GIVEN** analyze_app has completed
-- **WHEN** client requests `webtest://abc123/analysis/flows.md`
-- **THEN** it SHALL return user flows as markdown with YAML frontmatter containing structured flow definitions
-
-#### Scenario: Tests are accessible
-
-- **GIVEN** generate_tests has completed
-- **WHEN** client requests `webtest://abc123/tests/tests.md`
-- **THEN** it SHALL return the test cases as markdown with YAML frontmatter containing structured test definitions
-
-#### Scenario: Test run report is accessible
-
-- **GIVEN** a test run has completed with runId "run-001"
-- **WHEN** client requests `webtest://abc123/runs/run-001/report.md`
-- **THEN** it SHALL return the test execution report as markdown with YAML frontmatter containing structured results
-
-#### Scenario: Test step snapshot is accessible
-
-- **GIVEN** a test step has captured evidence
-- **WHEN** client requests `webtest://abc123/runs/run-001/steps/1/snapshot.md`
-- **THEN** it SHALL return the accessibility snapshot as formatted markdown with YAML frontmatter
-
-### Requirement: Resource Template Registration
-
-The system SHALL register resource templates with the MCP server for discovery, using markdown extensions for all index and report resources.
-
-#### Scenario: Templates are listed on resources/list
-
-- **GIVEN** a client calls `resources/list`
-- **WHEN** the response is returned
-- **THEN** it SHALL include templates for:
-  - `webtest://{analysisId}/index.md` (Analysis index)
-  - `webtest://{analysisId}/crawls/{crawlId}/index.md` (Crawl index)
-  - `webtest://{analysisId}/crawls/{crawlId}/checkpoint.md` (Crawl checkpoint)
-  - `webtest://{analysisId}/crawls/{crawlId}/pages/{pageId}/snapshot.md` (Page snapshot)
-  - `webtest://{analysisId}/crawls/{crawlId}/pages/{pageId}/screenshot.png` (Page screenshot)
-  - `webtest://{analysisId}/crawls/{crawlId}/pages/{pageId}/dom.html` (Page DOM)
-  - `webtest://{analysisId}/analysis/app-analysis.md` (Analysis report)
-  - `webtest://{analysisId}/analysis/flows.md` (User flows)
-  - `webtest://{analysisId}/tests/tests.md` (Test definitions)
-  - `webtest://{analysisId}/runs/{runId}/report.md` (Test run report)
-  - `webtest://{analysisId}/runs/{runId}/steps/{stepNumber}/snapshot.md` (Step snapshot)
-  - `webtest://{analysisId}/runs/{runId}/steps/{stepNumber}/screenshot.png` (Step screenshot)
-
-### Requirement: Resource Content Types
-
-The system SHALL return appropriate MIME types for different artifact types.
-
-#### Scenario: Markdown resources have correct type
-
-- **GIVEN** client reads a `.md` resource
-- **WHEN** response is returned
-- **THEN** mimeType SHALL be `text/markdown`
-
-#### Scenario: Screenshot resources have correct type
-
-- **GIVEN** client reads a `.png` resource
-- **WHEN** response is returned
-- **THEN** mimeType SHALL be `image/png`
-- **AND** content SHALL be base64 encoded
-
-#### Scenario: HTML resources have correct type
-
-- **GIVEN** client reads a `.html` resource
-- **WHEN** response is returned
-- **THEN** mimeType SHALL be `text/html`
-

package/openspec/changes/archive/2025-12-18-use-markdown-artifacts/specs/webtest-tools/spec.md

@@ -1,121 +0,0 @@
-# webtest-tools Spec Delta
-
-## MODIFIED Requirements
-
-### Requirement: webtest_init Tool
-
-The system SHALL provide a `webtest_init` tool that initializes an analysis workspace for a target URL and focus, storing metadata in markdown format.
-
-#### Scenario: Start analysis with valid URL
-
-- **GIVEN** the tool is called with a valid URL and focus
-- **WHEN** execution completes
-- **THEN** it SHALL generate a unique `analysisId`
-- **AND** create workspace directories
-- **AND** write initial `index.md` metadata with YAML frontmatter
-- **AND** return `{ analysisId, workspaceRootPath, workspaceRootUri, statusUri }` where statusUri points to `index.md`
-
-### Requirement: webtest_crawl_app Tool
-
-The system SHALL provide a `webtest_crawl_app` tool that dynamically explores a web application, storing all crawl artifacts in markdown format.
-
-#### Scenario: Crawl captures artifacts at each checkpoint
-
-- **GIVEN** a crawl iteration completes an action
-- **WHEN** state is captured
-- **THEN** it SHALL call Playwright MCP `browser_snapshot` for accessibility tree
-- **AND** call `browser_take_screenshot` for visual evidence
-- **AND** optionally extract HTML DOM
-- **AND** store snapshot as `snapshot.md` with formatted accessibility tree and YAML frontmatter
-- **AND** store screenshot as PNG
-- **AND** store DOM as HTML
-
-#### Scenario: Crawl outputs complete results
-
-- **GIVEN** crawl has finalized
-- **WHEN** output is returned
-- **THEN** it SHALL include `crawlId`, `crawlIndexFilePath`, `crawlIndexUri` (pointing to `index.md`), `pages[]`, `summaryUri`
-
-### Requirement: Crawl Checkpointing
-
-The system SHALL implement checkpointing during crawl using markdown format to enable resumption and provide human-readable partial results on failure.
-
-#### Scenario: Checkpoint is written periodically
-
-- **GIVEN** a crawl is in progress
-- **WHEN** N steps have completed (configurable, default 5)
-- **THEN** it SHALL write a checkpoint to `webtest://{analysisId}/crawls/{crawlId}/checkpoint.md`
-- **AND** the checkpoint SHALL be markdown with YAML frontmatter including: current step, visited pages, action history, goal progress
-- **AND** the checkpoint body SHALL contain human-readable progress summary
-
-#### Scenario: Crawl can resume from checkpoint
-
-- **GIVEN** a crawl was interrupted (cancelled, error, timeout)
-- **AND** a checkpoint exists as `checkpoint.md`
-- **WHEN** `webtest_crawl_app` is called with `resume: true`
-- **THEN** it SHALL parse the checkpoint YAML frontmatter
-- **AND** continue from the last recorded state
-
-### Requirement: webtest_analyze_app Tool
-
-The system SHALL provide a `webtest_analyze_app` tool that reverse-engineers application structure from crawl data, outputting all results in markdown format.
-
-#### Scenario: Analyze app writes markdown report
-
-- **GIVEN** analysis is complete
-- **WHEN** output is generated
-- **THEN** it SHALL write `app-analysis.md` resource to workspace
-
-#### Scenario: Analyze app writes flows
-
-- **GIVEN** analysis is complete
-- **WHEN** output is generated
-- **THEN** it SHALL write `flows.md` resource to workspace as markdown with YAML frontmatter containing structured flow definitions
-
-#### Scenario: Analyze app outputs URIs
-
-- **GIVEN** analysis is complete
-- **WHEN** tool returns
-- **THEN** it SHALL include `appAnalysisFilePath`, `appAnalysisUri`, `flowsFilePath`, `flowsUri` all pointing to `.md` files
-
-### Requirement: webtest_generate_tests Tool
-
-The system SHALL provide a `webtest_generate_tests` tool that produces test cases from application analysis in a single markdown format.
-
-#### Scenario: Generate tests outputs structured format
-
-- **GIVEN** test generation completes
-- **WHEN** results are written
-- **THEN** it SHALL produce `tests.md` with human-readable format AND YAML frontmatter containing structured test definitions
-- **AND** there SHALL NOT be a separate `tests.json` file
-
-#### Scenario: Generate tests outputs URIs
-
-- **GIVEN** generation is complete
-- **WHEN** tool returns
-- **THEN** it SHALL include `testsFilePath` and `testsUri` pointing to `tests.md`
-
-### Requirement: webtest_run_tests Tool
-
-The system SHALL provide a `webtest_run_tests` tool that executes a test case with evidence capture, storing all results in markdown format.
-
-#### Scenario: Run test case captures evidence
-
-- **GIVEN** a step is executed
-- **WHEN** evidence is captured
-- **THEN** it SHALL take screenshot after action (stored as PNG)
-- **AND** capture accessibility snapshot (stored as `snapshot.md` with formatted tree and YAML frontmatter)
-- **AND** store with step identifier
-
-#### Scenario: Run test case outputs report
-
-- **GIVEN** test execution completes
-- **WHEN** output is generated
-- **THEN** it SHALL write `report.md` with pass/fail summary, step details, evidence links, and YAML frontmatter containing structured run data
-- **AND** there SHALL NOT be a separate `index.json` or `artifacts.json` file
-
-#### Scenario: Run test case returns URIs
-
-- **GIVEN** execution is complete
-- **WHEN** tool returns
-- **THEN** it SHALL include `testRunId`, `reportFilePath`, `reportUri` pointing to `report.md`

package/openspec/changes/archive/2025-12-18-use-markdown-artifacts/tasks.md

@@ -1,133 +0,0 @@
-# Tasks: Use Markdown for All Artifacts
-
-## Prerequisites
-
-- [x] **Add gray-matter dependency** - Install `gray-matter` package for YAML frontmatter parsing
-  - `npm install gray-matter`
-  - Add to package.json dependencies
-
-## Phase 1: Core Utilities
-
-- [x] **Create markdown utilities module** - `src/workspace/markdown.ts`
-  - Implement `writeMarkdownWithFrontmatter(path, frontmatter, content)`
-  - Implement `readMarkdownFrontmatter<T>(path)` returning parsed YAML data
-  - Implement `formatAccessibilityTree(snapshot)` for human-readable tree output
-  - Add unit tests for utilities
-
-## Phase 2: Workspace Index
-
-- [x] **Update workspace index to markdown** - `src/workspace/index.ts`
-  - Change `createWorkspace` to write `index.md` instead of `index.json`
-  - Update `readWorkspaceIndex` to parse from markdown frontmatter
-  - Update `updateWorkspaceIndex` to preserve markdown body while updating frontmatter
-  - Generate human-readable workspace summary in markdown body
-
-- [x] **Update workspace types** - `src/workspace/types.ts`
-  - Update `WorkspaceIndex` type if needed for markdown-specific fields
-
-## Phase 3: Crawl Artifacts
-
-- [x] **Update crawl index to markdown** - `src/workspace/index.ts`
-  - Change `createCrawl` to write `index.md`
-  - Update `readCrawlIndex` to parse from markdown frontmatter
-  - Update `updateCrawlIndex` to regenerate markdown body
-  - Include formatted action history and budget status in body
-
-- [x] **Update checkpoint to markdown** - `src/workspace/index.ts`
-  - Change `saveCheckpoint` to write `checkpoint.md`
-  - Update `loadCheckpoint` to parse from markdown frontmatter
-  - Include human-readable progress summary in body
-
-- [x] **Update page snapshot to markdown** - `src/workspace/index.ts`
-  - Change `savePage` to write `snapshot.md` instead of `snapshot.json`
-  - Format accessibility tree as indented text tree in body
-  - Keep full structured snapshot in YAML frontmatter
-  - Update return paths/URIs to `.md` extension
-
-## Phase 4: Analysis Artifacts
-
-- [x] **Update flows to markdown** - `src/tools/webtest/analyze-app.ts`
-  - Change `saveAnalysis` to write `flows.md` instead of `flows.json`
-  - Format flows as readable markdown sections
-  - Include structured flow data in YAML frontmatter
-  - Update return paths/URIs
-
-- [x] **Update analyze-app tool outputs** - `src/tools/webtest/analyze-app.ts`
-  - Update `flowsFilePath` and `flowsUri` to point to `flows.md`
-
-## Phase 5: Test Artifacts
-
-- [x] **Consolidate test definitions** - `src/tools/webtest/generate-tests.ts`
-  - Remove separate `tests.json` generation
-  - Update `saveTests` to write only `tests.md` with YAML frontmatter
-  - Include full test case structures in frontmatter
-  - Keep human-readable test documentation in body
-  - Update return paths/URIs (remove `testsFilePath` for JSON)
-
-- [x] **Update generate-tests tool outputs** - `src/tools/webtest/generate-tests.ts`
-  - Return single `testsUri` pointing to `tests.md`
-  - Update `testsFilePath` to point to `tests.md`
-
-## Phase 6: Test Run Artifacts
-
-- [x] **Update test run index to report** - `src/workspace/index.ts`
-  - Change `createTestRun` to write `report.md` instead of `index.json`
-  - Update `readTestRunIndex` to parse from markdown frontmatter
-  - Update `updateTestRunIndex` to regenerate markdown body
-
-- [x] **Update test step snapshots** - `src/workspace/index.ts`
-  - Change `saveTestStepEvidence` to write `snapshot.md`
-  - Format accessibility tree in body
-  - Update return paths/URIs
-
-- [x] **Update run-test-case tool outputs** - `src/tools/webtest/run-test-case.ts`
-  - Update `reportUri` to point to `report.md`
-  - Update step evidence URIs to point to `snapshot.md`
-
-## Phase 7: Resource Manager
-
-- [x] **Update resource URI mappings** - `src/resources/index.ts`
-  - Update all `.json` paths to `.md` in URI templates
-  - Handle markdown MIME type for new paths
-  - Update resource listing to discover `.md` files
-
-- [x] **Update resource reading** - `src/resources/index.ts`
-  - Return raw markdown content for `.md` resources
-  - Update MIME type to `text/markdown`
-
-## Phase 8: Integration & Testing
-
-- [x] **Update existing tests** - Test files in `src/` or `tests/`
-  - Update file path expectations from `.json` to `.md`
-  - Update content parsing to use frontmatter extraction
-  - Add tests for markdown format validity
-
-- [x] **Manual integration test**
-  - Run full workflow: init → crawl → analyze → generate tests → run tests
-  - Verify all workspace files are `.md` (except screenshots and DOM)
-  - Verify files are readable without tooling
-  - Verify programmatic access via frontmatter works
-
-## Phase 9: Documentation
-
-- [x] **Update CLAUDE.md or README** (if exists)
-  - Document new artifact format
-  - Provide examples of frontmatter structure
-  - Note breaking change for existing workspaces
-
-## Verification Checklist
-
-After implementation, verify:
-
-- [x] `index.md` at workspace root with valid frontmatter
-- [x] `crawls/{id}/index.md` with action history
-- [x] `crawls/{id}/checkpoint.md` with progress state
-- [x] `crawls/{id}/pages/{id}/snapshot.md` with formatted tree
-- [x] `analysis/app-analysis.md` (unchanged)
-- [x] `analysis/flows.md` with flow definitions
-- [x] `tests/tests.md` with all test data (no tests.json)
-- [x] `runs/{id}/report.md` with execution results
-- [x] `runs/{id}/steps/{n}/snapshot.md` with formatted tree
-- [x] All tool outputs reference `.md` paths
-- [x] All resource URIs use `.md` extensions
-- [x] Workspace is browsable and human-readable

package/openspec/changes/extract-prompts-to-markdown/design.md

@@ -1,86 +0,0 @@
-## Context
-
-Prompts are currently embedded as template literal strings in TypeScript files. This change extracts them to markdown files for better maintainability while preserving runtime variable interpolation.
-
-## Goals / Non-Goals
-
-**Goals:**
-- Make prompts easy to read, edit, and review as standalone files
-- Preserve `${varName}` syntax for variable substitution
-- Maintain type safety for required variables
-- Keep runtime performance acceptable (load once, cache)
-
-**Non-Goals:**
-- Complex templating features (loops, conditionals) - keep it simple
-- Hot-reloading of templates in production
-- User-configurable prompts at runtime
-
-## Decisions
-
-### Decision: Use simple `${varName}` interpolation
-
-Keep the existing JavaScript template literal syntax. The loader will:
-1. Read the markdown file content
-2. Replace `${varName}` patterns with provided values
-3. Return the interpolated string
-
-**Alternatives considered:**
-- Mustache/Handlebars: Adds dependency, overkill for simple substitution
-- Tagged template literals: Would require keeping prompts in JS files
-- JSON with placeholders: Less readable for long-form content
-
-### Decision: Directory structure
-
-```
-src/prompts/
-├── templates/
-│   ├── mcp/           # User-facing MCP prompts
-│   │   ├── webtest-start.md
-│   │   ├── webtest-crawl.md
-│   │   └── ...
-│   └── sampling/      # Internal AI sampling prompts
-│       ├── system-prefix.md
-│       ├── crawl-action.md
-│       └── ...
-├── loader.ts          # Template loading utility
-├── index.ts           # MCP prompt definitions (uses loader)
-└── index.test.ts      # Tests
-```
-
-**Rationale:** Separating `mcp/` and `sampling/` makes it clear which prompts are user-facing vs internal.
-
-### Decision: Synchronous file loading with caching
-
-Load templates synchronously at module initialization and cache them. This:
-- Avoids async complexity in prompt builders
-- Ensures templates are available immediately
-- Is acceptable since templates are static assets
-
-### Decision: TypeScript interface for variables
-
-Each prompt template will have a corresponding TypeScript type defining its variables:
-
-```typescript
-interface CrawlActionVars {
-  goal: string;
-  currentUrl: string;
-  pageSnapshot: string;
-  actionHistory: string;
-  allowedDomains: string;
-  // ... optional vars
-}
-```
-
-This maintains type safety when calling the interpolation function.
-
-## Risks / Trade-offs
-
-| Risk | Mitigation |
-|------|------------|
-| File system access at startup | Templates are small, load is fast |
-| Missing variable errors at runtime | Type definitions + tests catch this |
-| Bundler complications | Ensure templates are included in build output |
-
-## Open Questions
-
-- None at this time

package/openspec/changes/extract-prompts-to-markdown/proposal.md

@@ -1,50 +0,0 @@
-# Change: Extract prompts to separate markdown files
-
-## Why
-
-Currently, all prompt text is embedded inline within TypeScript files (`src/prompts/index.ts` and `src/sampling/prompts.ts`). This makes prompts difficult to maintain, review, and iterate on since:
-- Prompt changes require modifying TypeScript code
-- Large multi-line strings with template literals are hard to read and edit
-- Non-developers cannot easily review or suggest prompt improvements
-- No syntax highlighting or markdown preview for prompt content
-
-## What Changes
-
-- **Extract MCP prompts** from `src/prompts/index.ts` to separate `.md` files
-- **Extract sampling prompts** from `src/sampling/prompts.ts` to separate `.md` files
-- Store all prompt templates in `src/prompts/templates/`
-- Keep `${varName}` interpolation syntax for dynamic values
-- Create a utility to load and interpolate markdown templates at runtime
-- Update existing code to use the template loader
-
-### Prompts to Extract
-
-**MCP Prompts (user-facing):**
-- `webtest-start.md`
-- `webtest-crawl.md`
-- `webtest-discover.md`
-- `webtest-generate-tests.md`
-- `webtest-run-test.md`
-- `webtest-full-workflow.md`
-
-**Sampling Prompts (internal AI prompts):**
-- `system-prefix.md` - Security-hardened system prompt
-- `page-content-wrapper.md` - Untrusted content delimiters
-- `crawl-action.md` - Crawl navigation decisions
-- `feature-discovery.md` - Discover app features/modules
-- `flow-discovery.md` - Discover user flows within features
-- `test-generation.md` - Generate test cases
-- `test-evaluation.md` - Evaluate test step results
-
-**Cleanup:**
-- Remove deprecated `buildAnalysisPrompt` function
-- Update `analyze-app.ts` to use `buildFeatureDiscoveryPrompt` instead
-
-## Impact
-
-- Affected specs: `webtest-prompts`
-- Affected code:
-  - `src/prompts/index.ts` - Refactor to load from markdown
-  - `src/sampling/prompts.ts` - Refactor to load from markdown
-  - New: `src/prompts/templates/` directory with `.md` files
-  - New: `src/prompts/loader.ts` - Template loading utility

package/openspec/changes/extract-prompts-to-markdown/specs/webtest-prompts/spec.md

@@ -1,74 +0,0 @@
-## ADDED Requirements
-
-### Requirement: Prompt Template Files
-
-The system SHALL store prompt text in separate markdown files under `src/prompts/templates/`.
-
-#### Scenario: MCP prompts stored in templates/mcp/
-
-- **GIVEN** a user-facing MCP prompt (e.g., `webtest-start`)
-- **WHEN** the prompt content is needed
-- **THEN** it SHALL be loaded from `src/prompts/templates/mcp/<prompt-name>.md`
-
-#### Scenario: Sampling prompts stored in templates/sampling/
-
-- **GIVEN** an internal sampling prompt (e.g., `crawl-action`)
-- **WHEN** the prompt content is needed
-- **THEN** it SHALL be loaded from `src/prompts/templates/sampling/<prompt-name>.md`
-
-### Requirement: Template Variable Interpolation
-
-The system SHALL support `${varName}` syntax for variable substitution in markdown templates.
-
-#### Scenario: Simple variable substitution
-
-- **GIVEN** a template containing `${url}` placeholder
-- **WHEN** the template is rendered with `{ url: "https://example.com" }`
-- **THEN** the output SHALL contain `https://example.com` in place of `${url}`
-
-#### Scenario: Multiple variables in template
-
-- **GIVEN** a template containing `${goal}`, `${currentUrl}`, and `${allowedDomains}`
-- **WHEN** the template is rendered with all variables provided
-- **THEN** all placeholders SHALL be replaced with their corresponding values
-
-#### Scenario: Preserve unmatched placeholders
-
-- **GIVEN** a template containing `${undefinedVar}`
-- **WHEN** the template is rendered without that variable
-- **THEN** the placeholder `${undefinedVar}` SHALL remain in the output unchanged
-- **AND** no error SHALL be thrown
-
-### Requirement: Template Loading and Caching
-
-The system SHALL load templates efficiently with appropriate caching.
-
-#### Scenario: Templates loaded at initialization
-
-- **GIVEN** the server starts
-- **WHEN** templates are first accessed
-- **THEN** they SHALL be loaded from the filesystem
-- **AND** cached for subsequent access
-
-#### Scenario: Cached templates returned on repeated access
-
-- **GIVEN** a template has been loaded previously
-- **WHEN** the same template is requested again
-- **THEN** the cached version SHALL be returned without re-reading the file
-
-### Requirement: Template File Format
-
-Markdown template files SHALL follow a consistent format for maintainability.
-
-#### Scenario: Template uses raw markdown content
-
-- **GIVEN** a prompt template file
-- **WHEN** the file is read
-- **THEN** the entire file content SHALL be used as the prompt text
-- **AND** no special parsing or metadata extraction is required
-
-#### Scenario: Template preserves whitespace and formatting
-
-- **GIVEN** a template with specific indentation and line breaks
-- **WHEN** the template is loaded and interpolated
-- **THEN** all whitespace and formatting SHALL be preserved exactly

package/openspec/changes/extract-prompts-to-markdown/tasks.md

@@ -1,40 +0,0 @@
-## 1. Template Infrastructure
-
-- [x] 1.1 Create `src/prompts/templates/` directory structure
-- [x] 1.2 Implement `src/prompts/loader.ts` with template loading and interpolation
-- [x] 1.3 Add tests for template loader (variable substitution, missing vars, file not found)
-
-## 2. Extract Sampling Prompts
-
-- [x] 2.1 Create `src/prompts/templates/sampling/system-prefix.md`
-- [x] 2.2 Create `src/prompts/templates/sampling/page-content-wrapper.md`
-- [x] 2.3 Create `src/prompts/templates/sampling/crawl-action.md`
-- [x] 2.4 Create `src/prompts/templates/sampling/feature-discovery.md`
-- [x] 2.5 Create `src/prompts/templates/sampling/flow-discovery.md`
-- [x] 2.6 Create `src/prompts/templates/sampling/test-generation.md`
-- [x] 2.7 Create `src/prompts/templates/sampling/test-evaluation.md`
-- [x] 2.8 Update `src/sampling/prompts.ts` to use template loader
-- [x] 2.9 Verify existing sampling tests pass
-
-## 3. Remove Deprecated Code
-
-- [x] 3.1 Remove `buildAnalysisPrompt` from `src/sampling/prompts.ts` (already removed)
-- [x] 3.2 Update `src/tools/webtest/analyze-app.ts` to use `buildFeatureDiscoveryPrompt` (already done)
-- [x] 3.3 Update or remove related tests in `analyze-app.test.ts` (already done)
-
-## 4. Extract MCP Prompts
-
-- [x] 4.1 Create `src/prompts/templates/mcp/webtest-start.md`
-- [x] 4.2 Create `src/prompts/templates/mcp/webtest-crawl.md`
-- [x] 4.3 Create `src/prompts/templates/mcp/webtest-discover.md`
-- [x] 4.4 Create `src/prompts/templates/mcp/webtest-discover-flows.md`
-- [x] 4.5 Create `src/prompts/templates/mcp/webtest-generate-tests.md`
-- [x] 4.6 Create `src/prompts/templates/mcp/webtest-run-test.md`
-- [x] 4.7 Create `src/prompts/templates/mcp/webtest-full-workflow.md`
-- [x] 4.8 Update `src/prompts/index.ts` to use template loader
-- [x] 4.9 Verify existing MCP prompt tests pass
-
-## 5. Validation
-
-- [x] 5.1 Run full test suite (`npm test`) - 448 tests pass
-- [x] 5.2 Run build (`npm run build`) - success