Major codebase cleanup and refactoring

.claude/settings.local.json

@@ -0,0 +1,25 @@
+{
+  "permissions": {
+    "allow": [
+      "Bash(git push:*)",
+      "WebFetch(domain:code.visualstudio.com)",
+      "mcp__too-many-cooks__lock",
+      "Bash(npm run compile:*)",
+      "Bash(node -e:*)",
+      "Bash(npx tsc:*)",
+      "Bash(npm run lint:*)",
+      "mcp__too-many-cooks__message",
+      "mcp__too-many-cooks__register",
+      "Bash(git add:*)",
+      "Bash(git commit:*)",
+      "Bash(find /Users/christianfindlay/Documents/Code/tmc/too-many-cooks -name test*.sh -o -name *test.sh)",
+      "mcp__too-many-cooks__plan",
+      "Bash(npm ci:*)",
+      "Bash(npm run:*)",
+      "Bash(npx cspell:*)",
+      "Bash(gh pr:*)",
+      "Bash(gh run:*)"
+    ]
+  },
+  "autoMemoryEnabled": false
+}

.claude/skills/ci-prep/SKILL.md

@@ -0,0 +1,62 @@
+---
+name: ci-prep
+description: Prepare the codebase for CI. Reads the CI workflow, builds a checklist, then loops through format/lint/build/test/coverage until every single check passes. Use before submitting a PR or when the user wants to ensure CI will pass.
+argument-hint: "[optional focus area]"
+allowed-tools: Read, Grep, Glob, Edit, Write, Bash
+---
+
+# CI Prep — Get the Codebase PR-Ready
+
+You MUST NOT STOP until every check passes and coverage threshold is met.
+
+## Step 1: Read the CI Pipeline and Build Your Checklist
+
+Read the CI workflow file:
+
+```bash
+cat .github/workflows/ci.yml
+```
+
+Parse EVERY step in the workflow. Extract the exact commands CI runs. Build yourself a numbered checklist of every check you need to pass. This is YOUR checklist — derived from the actual CI config, not from assumptions. The CI pipeline changes over time so you MUST read it fresh and build your list from what you find.
+
+## Step 2: Coordinate with Other Agents
+
+You are likely working alongside other agents who are editing files concurrently. Before making changes:
+
+1. Check TMC status and messages for active agents and locked files
+2. Do NOT edit files that are locked by other agents
+3. Lock files before editing them yourself
+4. Communicate what you are doing via TMC broadcasts
+5. After each fix cycle, check TMC again — another agent may have broken something
+
+## Step 3: The Loop
+
+Run through your checklist from Step 1 in order. For each check:
+
+1. Run the exact command from CI
+2. If it passes, move to the next check
+3. If it fails, FIX IT. Do NOT suppress warnings, ignore errors, remove assertions, or lower thresholds. Fix the actual code.
+4. Re-run that check to confirm the fix works
+5. Move to the next check
+
+When you reach the end of the checklist, GO BACK TO THE START AND RUN THE ENTIRE CHECKLIST AGAIN. Other agents are working concurrently and may have broken something you already fixed. A fix for one check may have broken an earlier check.
+
+**Keep looping through the full checklist until you get a COMPLETE CLEAN RUN with ZERO failures from start to finish.** One clean pass is not enough if you fixed anything during that pass — you need a clean pass where NOTHING needed fixing.
+
+Do NOT stop after one loop. Do NOT stop after two loops. Keep going until a full pass completes with every single check green on the first try.
+
+## Step 4: Final Coordination
+
+1. Broadcast on TMC that CI prep is complete and all checks pass
+2. Release any locks you hold
+3. Report the final status to the user with the output of each passing check
+
+## Rules
+
+- NEVER stop with failing checks. Loop until everything is green.
+- NEVER suppress lint warnings, skip tests, or lower coverage thresholds.
+- NEVER remove assertions to make tests pass.
+- Fix the CODE, not the checks.
+- If you are stuck on a failure after 3 attempts on the same issue, ask the user for help. Do NOT silently give up.
+- Always coordinate with other agents via TMC. Check for messages regularly.
+- Leave the codebase in a state that will pass CI on the first try.

.claude/skills/fix-bug/SKILL.md

@@ -0,0 +1,66 @@
+---
+name: fix-bug
+description: Fix a bug using test-driven development. Use when the user reports a bug, describes unexpected behavior, wants to fix a defect, or says something is broken. Enforces a strict test-first workflow where a failing test must be written and verified before any fix is attempted.
+argument-hint: "[bug description]"
+allowed-tools: Read, Grep, Glob, Edit, Write, Bash
+---
+
+# Bug Fix Skill — Test-First Workflow
+
+You MUST follow this exact workflow. Do NOT skip steps. Do NOT fix the bug before writing a failing test.
+
+## Step 1: Understand the Bug
+
+- Read the bug description: $ARGUMENTS
+- Investigate the codebase to understand the relevant code
+- Identify the root cause (or narrow down candidates)
+- Summarize your understanding of the bug to the user before proceeding
+
+## Step 2: Write a Failing Test
+
+- Write a test that **directly exercises the buggy behavior**
+- The test must assert the **correct/expected** behavior — so it FAILS against the current broken code
+- The test name should clearly describe the bug (e.g., `test_orange_color_not_applied_to_head`)
+- Use the project's existing test framework and conventions
+
+## Step 3: Run the Test — Confirm It FAILS
+
+- Run ONLY the new test (not the full suite)
+- **Verify the test FAILS** and that it fails **because of the bug**, not for some other reason (typo, import error, wrong selector, etc.)
+- If the test passes: your test does not capture the bug. Go back to Step 2
+- If the test fails for the wrong reason: fix the test, not the code. Go back to Step 2
+- **Repeat until the test fails specifically because of the bug**
+
+## Step 4: Show Failure to User
+
+- Show the user the test code and the failure output
+- Explicitly ask: "This test fails because of the bug. Can you confirm this captures the issue before I fix it?"
+- **STOP and WAIT for user acknowledgment before proceeding**
+- Do NOT continue to Step 5 until the user confirms
+
+## Step 5: Fix the Bug
+
+- Make the **minimum change** needed to fix the bug
+- Do not refactor, clean up, or "improve" surrounding code
+- Do not change the test
+
+## Step 6: Run the Test — Confirm It PASSES
+
+- Run the new test again
+- **Verify it PASSES**
+- If it fails: go back to Step 5 and adjust the fix
+- **Repeat until the test passes**
+
+## Step 7: Run the Full Test Suite
+
+- Run ALL tests to make sure nothing else broke
+- If other tests fail: fix the regression without breaking the new test
+- Report the final result to the user
+
+## Rules
+
+- NEVER fix the bug before the failing test is written and confirmed
+- NEVER skip asking the user to acknowledge the test failure
+- NEVER modify the test to make it pass — modify the source code
+- If you cannot write a test for the bug, explain why and ask the user how to proceed
+- Keep the fix minimal — one bug, one fix, one test

.github/workflows/ci.yml

@@ -17,6 +17,9 @@ jobs:
 
       - run: npm ci
 
+      - name: Format check
+        run: npm run format:check
+
       - name: Lint
         run: npm run lint
 
@@ -31,8 +34,11 @@ jobs:
       - name: Unit tests
         run: npm run test:unit
 
-      - name: E2E tests
-        run: xvfb-run -a npm run test:e2e
+      - name: E2E tests with coverage
+        run: xvfb-run -a npx vscode-test --coverage --grep @exclude-ci --invert
+
+      - name: Coverage threshold (90%)
+        run: npm run coverage:check
 
   website:
     runs-on: ubuntu-latest

.gitignore

@@ -32,3 +32,5 @@ src/test/fixtures/workspace/.vscode/tasktree.json
 website/_site/
 
 .commandtree/
+
+logs/

.mcp.json

@@ -0,0 +1,3 @@
+{
+  "mcpServers": {}
+}

.prettierrc

@@ -0,0 +1,7 @@
+{
+  "semi": true,
+  "singleQuote": false,
+  "trailingComma": "es5",
+  "tabWidth": 2,
+  "printWidth": 120
+}

.vscode-test.mjs

@@ -14,24 +14,31 @@ cpSync('./src/test/fixtures/workspace', testWorkspace, { recursive: true });
 const userDataDir = resolve(__dirname, '.vscode-test/user-data');
 
 export default defineConfig({
-    files: ['out/test/e2e/**/*.test.js', 'out/test/providers/**/*.test.js'],
-    version: 'stable',
-    workspaceFolder: testWorkspace,
-    extensionDevelopmentPath: './',
-    mocha: {
-        ui: 'tdd',
-        timeout: 60000,
-        color: true,
-        slow: 10000
-    },
-    launchArgs: [
-        '--disable-gpu',
-        '--user-data-dir', userDataDir
-    ],
+    tests: [{
+        files: ['out/test/e2e/**/*.test.js', 'out/test/providers/**/*.test.js'],
+        version: 'stable',
+        workspaceFolder: testWorkspace,
+        extensionDevelopmentPath: './',
+        mocha: {
+            ui: 'tdd',
+            timeout: 60000,
+            color: true,
+            slow: 10000
+        },
+        launchArgs: [
+            '--disable-gpu',
+            '--user-data-dir', userDataDir
+        ]
+    }],
     coverage: {
         include: ['out/**/*.js'],
-        exclude: ['out/test/**/*.js'],
-        reporter: ['text', 'lcov', 'html'],
+        exclude: [
+            'out/test/**/*.js',
+            'out/semantic/summariser.js',       // requires Copilot auth, not available in CI
+            'out/semantic/summaryPipeline.js',   // requires Copilot auth, not available in CI
+            'out/semantic/vscodeAdapters.js',    // requires Copilot auth, not available in CI
+        ],
+        reporter: ['text', 'lcov', 'html', 'json-summary'],
         output: './coverage'
     }
 });

Agents.md

@@ -11,9 +11,10 @@ You are working with many other agents. Make sure there is effective cooperation
 
 - **Zero duplication - TOP PRIORITY** - Always search for existing code before adding. Move; don't copy files. Add assertions to tests rather than duplicating tests. AIM FOR LESS CODE!
 - **No string literals** - Named constants only, and it ONE location
+- DO NOT USE GIT
 - **Functional style** - Prefer pure functions, avoid classes where possible
 - **No suppressing warnings** - Fix them properly
-- **No REGEX** It is absolutely ⛔️ illegal
+- **No REGEX** It is absolutely ⛔️ illegal, and no text matching in general
 - **Don't run long runnings tasks** like docker builds, tests. Ask the user to do it!!
 - **Expressions over assignments** - Prefer const and immutable patterns
 - **Named parameters** - Use object params for functions with 3+ args
@@ -23,6 +24,8 @@ You are working with many other agents. Make sure there is effective cooperation
 
 ### Typescript
 - **TypeScript strict mode** - No `any`, no implicit types, turn all lints up to error
+- **Regularly run the linter** - Fix lint errors IMMEDIATELY
+- **Decouple providers from the VSCODE SDK** - No vscode sdk use within the providers
 - **Ignoring lints = ⛔️ illegal** - Fix violations immediately
 - **No throwing** - Only return `Result<T,E>`
 
@@ -36,7 +39,6 @@ You are working with many other agents. Make sure there is effective cooperation
 
 #### Rules
 - **Prefer e2e tests over unit tests** - only unit tests for isolating bugs
-- DO NOT USE GIT
 - Separate e2e tests from unit tests by file. They should not be in the same file together.
 - Prefer adding assertions to existing tests rather than adding new tests
 - Test files in `src/test/suite/*.test.ts`
@@ -96,8 +98,21 @@ assert.ok(true, 'Command ran');
 
 ## Critical Docs
 
+### Vscode SDK
 [VSCode Extension API](https://code.visualstudio.com/api/)
-[SCode Extension Testing API](https://code.visualstudio.com/api/extension-guides/testing)
+[VSCode Extension Testing API](https://code.visualstudio.com/api/extension-guides/testing)
+[VSCODE Language Model API](https://code.visualstudio.com/api/extension-guides/ai/language-model)
+[Language Model Tool API](https://code.visualstudio.com/api/extension-guides/ai/tools)
+[AI extensibility in VS Cod](https://code.visualstudio.com/api/extension-guides/ai/ai-extensibility-overview)
+[AI language models in VS Code](https://code.visualstudio.com/docs/copilot/customization/language-models)
+
+### Website
+
+https://developers.google.com/search/blog/2025/05/succeeding-in-ai-search
+https://developers.google.com/search/docs/fundamentals/seo-starter-guide
+
+https://studiohawk.com.au/blog/how-to-optimise-ai-overviews/
+https://about.ads.microsoft.com/en/blog/post/october-2025/optimizing-your-content-for-inclusion-in-ai-search-answers
 
 ## Project Structure
 
@@ -110,11 +125,25 @@ CommandTree/
 │   │   └── TagConfig.ts      # Tag configuration from commandtree.json
 │   ├── discovery/
 │   │   ├── index.ts          # Discovery orchestration
-│   │   ├── shell.ts          # Shell script discovery
-│   │   ├── npm.ts            # NPM script discovery
-│   │   ├── make.ts           # Makefile target discovery
-│   │   ├── launch.ts         # launch.json discovery
-│   │   └── tasks.ts          # tasks.json discovery
+│   │   ├── shell.ts          # Shell scripts (.sh, .bash, .zsh)
+│   │   ├── npm.ts            # NPM scripts (package.json)
+│   │   ├── make.ts           # Makefile targets
+│   │   ├── launch.ts         # VS Code launch configs
+│   │   ├── tasks.ts          # VS Code tasks
+│   │   ├── python.ts         # Python scripts (.py)
+│   │   ├── powershell.ts     # PowerShell scripts (.ps1)
+│   │   ├── gradle.ts         # Gradle tasks
+│   │   ├── cargo.ts          # Cargo (Rust) tasks
+│   │   ├── maven.ts          # Maven goals (pom.xml)
+│   │   ├── ant.ts            # Ant targets (build.xml)
+│   │   ├── just.ts           # Just recipes (justfile)
+│   │   ├── taskfile.ts       # Taskfile tasks (Taskfile.yml)
+│   │   ├── deno.ts           # Deno tasks (deno.json)
+│   │   ├── rake.ts           # Rake tasks (Rakefile)
+│   │   ├── composer.ts       # Composer scripts (composer.json)
+│   │   ├── docker.ts         # Docker Compose services
+│   │   ├── dotnet.ts         # .NET projects (.csproj)
+│   │   └── markdown.ts       # Markdown files (.md)
 │   ├── models/
 │   │   └── TaskItem.ts       # Task data model and TreeItem
 │   ├── runners/

Claude.md

@@ -14,15 +14,15 @@ You are working with many other agents. Make sure there is effective cooperation
 - DO NOT USE GIT
 - **Functional style** - Prefer pure functions, avoid classes where possible
 - **No suppressing warnings** - Fix them properly
-- **No REGEX** It is absolutely ⛔️ illegal, and no text matching in general
-- **Don't run long runnings tasks** like docker builds, tests. Ask the user to do it!!
+- Text matching (including Regex) is illegal. Use a proper parser/treesitter. If text matching is absolutely necessary, prefer Regex
 - **Expressions over assignments** - Prefer const and immutable patterns
 - **Named parameters** - Use object params for functions with 3+ args
 - **Keep files under 450 LOC and functions under 20 LOC**
 - **No commented-out code** - Delete it
 - **No placeholders** - If incomplete, leave LOUD compilation error with TODO
 
 ### Typescript
+- **CENTRALIZE global state** Keep it in one type/file.
 - **TypeScript strict mode** - No `any`, no implicit types, turn all lints up to error
 - **Regularly run the linter** - Fix lint errors IMMEDIATELY
 - **Decouple providers from the VSCODE SDK** - No vscode sdk use within the providers
@@ -40,6 +40,8 @@ You are working with many other agents. Make sure there is effective cooperation
 #### Rules
 - **Prefer e2e tests over unit tests** - only unit tests for isolating bugs
 - Separate e2e tests from unit tests by file. They should not be in the same file together.
+- Tests must prove USER INTERACTIONS work
+- E2E tests should have multiple user interactions each and loads of assertions
 - Prefer adding assertions to existing tests rather than adding new tests
 - Test files in `src/test/suite/*.test.ts`
 - Run tests: `npm test`