Hardening: audit, accuracy fixes, isolation and credential fixes, tests, CI

.editorconfig

@@ -0,0 +1,12 @@
+root = true
+
+[*]
+charset = utf-8
+end_of_line = lf
+insert_final_newline = true
+indent_style = space
+indent_size = 2
+trim_trailing_whitespace = true
+
+[*.md]
+trim_trailing_whitespace = false

.env.example

@@ -0,0 +1,8 @@
+# Copy to .env (gitignored). Loaded automatically by `pnpm harness`.
+# Only needed for the GitHub experiment. Use fine-grained PATs scoped to a sandbox org.
+GITHUB_SANDBOX_OWNER=
+GITHUB_CONTROLLER_TOKEN=
+GITHUB_AGENT_TOKEN=
+# Tier 2 (github-rw) runs: pass this as GITHUB_AGENT_TOKEN for that command (see README).
+GITHUB_AGENT_TOKEN_RW=
+# GITHUB_HOST=api.github.com

.github/ISSUE_TEMPLATE/bug_report.md

@@ -0,0 +1,15 @@
+---
+name: Bug report
+about: Something in the harness doesn't work as documented
+labels: bug
+---
+
+**What happened**
+
+**What you expected**
+
+**How to reproduce** (command, experiment, arm, task)
+
+**Environment**: OS, `node -v`, `pnpm -v`, `claude --version`
+
+Don't paste tokens or unredacted transcripts. Report security issues privately (see SECURITY.md).

.github/ISSUE_TEMPLATE/config.yml

@@ -0,0 +1,5 @@
+blank_issues_enabled: true
+contact_links:
+  - name: Report a security vulnerability
+    url: https://github.com/chief-builder/cli-vs-mcp/security/advisories/new
+    about: Please report security issues privately.

.github/ISSUE_TEMPLATE/feature_request.md

@@ -0,0 +1,11 @@
+---
+name: Feature or experiment idea
+about: Suggest a new experiment, task, or harness feature
+labels: enhancement
+---
+
+**What you'd like to measure or change**
+
+**Why it matters**
+
+**Tool surfaces involved** (CLI / MCP server, versions)

.github/codeql/codeql-config.yml

@@ -0,0 +1,3 @@
+# Committed run artifacts include scripts written by the agent under test; they are data, not project code.
+paths-ignore:
+  - experiments/*/runs/**

.github/dependabot.yml

@@ -0,0 +1,22 @@
+version: 2
+updates:
+  - package-ecosystem: npm
+    directory: /
+    schedule:
+      interval: weekly
+    groups:
+      dev-tooling:
+        dependency-type: development
+    ignore:
+      # Experiment subject: upgrading changes what is measured. Bump together with a new run.
+      - dependency-name: '@playwright/cli'
+      # typescript-eslint does not support TypeScript 7 yet.
+      - dependency-name: typescript
+        update-types: ['version-update:semver-major']
+      # Keep in step with the pinned Node major (.nvmrc).
+      - dependency-name: '@types/node'
+        update-types: ['version-update:semver-major']
+  - package-ecosystem: github-actions
+    directory: /
+    schedule:
+      interval: weekly

.github/pull_request_template.md

@@ -0,0 +1,8 @@
+## What and why
+
+## How it was tested
+
+- [ ] `pnpm check` passes
+- [ ] `pnpm harness verify-arms` run (if arm tools or isolation changed)
+- [ ] Run artifacts redacted (`pnpm harness redact-artifacts`) and reviewed (if runs are committed)
+- [ ] Docs updated if claims, numbers, or commands changed

.github/secret_scanning.yml

@@ -0,0 +1,3 @@
+# Token-shaped strings under test/ are fake fixtures for redaction and env-scrub tests.
+paths-ignore:
+  - 'test/**'

.github/workflows/ci.yml

@@ -0,0 +1,58 @@
+name: CI
+
+on:
+  push:
+    branches: [main]
+  pull_request:
+
+permissions:
+  contents: read
+
+concurrency:
+  group: ${{ github.workflow }}-${{ github.ref }}
+  cancel-in-progress: true
+
+jobs:
+  check:
+    name: Lint, typecheck, test
+    runs-on: ubuntu-latest
+    timeout-minutes: 15
+    steps:
+      - uses: actions/checkout@3d3c42e5aac5ba805825da76410c181273ba90b1 # v7.0.1
+        with:
+          persist-credentials: false
+      - uses: pnpm/action-setup@ea17c68df8912ef543352723c149a84f56e3d413 # v6.1.0
+      - uses: actions/setup-node@820762786026740c76f36085b0efc47a31fe5020 # v7.0.0
+        with:
+          node-version-file: .nvmrc
+          cache: pnpm
+      - run: pnpm install --frozen-lockfile
+      - run: pnpm lint
+      - run: pnpm format:check
+      - run: pnpm typecheck
+      - name: Test with coverage
+        run: pnpm test:coverage
+      - name: Coverage summary
+        if: always()
+        run: |
+          node -e '
+            const s = require("./coverage/coverage-summary.json").total;
+            const row = k => `| ${k} | ${s[k].pct}% (${s[k].covered}/${s[k].total}) |`;
+            console.log(["| Metric | Coverage |", "|---|---|", ...["lines","statements","functions","branches"].map(row)].join("\n"));
+          ' >> "$GITHUB_STEP_SUMMARY"
+      # There is no compiled artifact (the harness runs from source via tsx); "build" checks the CLI loads.
+      - name: Build (CLI smoke)
+        run: pnpm harness --help
+
+  links:
+    name: Link check
+    runs-on: ubuntu-latest
+    timeout-minutes: 10
+    steps:
+      - uses: actions/checkout@3d3c42e5aac5ba805825da76410c181273ba90b1 # v7.0.1
+        with:
+          persist-credentials: false
+      - uses: lycheeverse/lychee-action@e7477775783ea5526144ba13e8db5eec57747ce8 # v2.9.0
+        with:
+          args: --config lychee.toml --no-progress './*.md' './docs/**/*.html' './.github/**/*.md'
+          fail: true

.github/workflows/codeql.yml

@@ -0,0 +1,40 @@
+name: CodeQL
+
+on:
+  push:
+    branches: [main]
+  pull_request:
+  schedule:
+    - cron: '23 4 * * 1'
+
+permissions:
+  contents: read
+
+concurrency:
+  group: ${{ github.workflow }}-${{ github.ref }}
+  cancel-in-progress: true
+
+jobs:
+  analyze:
+    name: Analyze (${{ matrix.language }})
+    runs-on: ubuntu-latest
+    timeout-minutes: 20
+    permissions:
+      contents: read
+      security-events: write
+    strategy:
+      fail-fast: false
+      matrix:
+        language: [javascript-typescript, actions]
+    steps:
+      - uses: actions/checkout@3d3c42e5aac5ba805825da76410c181273ba90b1 # v7.0.1
+        with:
+          persist-credentials: false
+      - uses: github/codeql-action/init@2892aa5e19bbd11bc0cff5427e3b750a04d9e3c2 # v4.38.2
+        with:
+          languages: ${{ matrix.language }}
+          build-mode: none
+          config-file: ./.github/codeql/codeql-config.yml
+      - uses: github/codeql-action/analyze@2892aa5e19bbd11bc0cff5427e3b750a04d9e3c2 # v4.38.2
+        with:
+          category: /language:${{ matrix.language }}

.gitignore

@@ -10,4 +10,14 @@ harness/dist/
 
 # Local secrets — never commit GitHub tokens
 .env
-.env.local
+.env.*
+!.env.example
+
+# Editors and local tool state
+.vscode/
+.idea/
+*.swp
+.claude/settings.local.json
+npm-debug.log*
+pnpm-debug.log*
+coverage/

.gitleaks.toml

@@ -0,0 +1,6 @@
+[extend]
+useDefault = true
+
+[allowlist]
+description = "Fake tokens used by redaction and env-scrub tests"
+paths = ['''^test/''']

.mcp.playwright.json

@@ -3,10 +3,7 @@
     "playwright": {
       "type": "stdio",
       "command": "npx",
-      "args": [
-        "-y",
-        "@playwright/mcp@0.0.75"
-      ],
+      "args": ["-y", "@playwright/mcp@0.0.75"],
       "env": {}
     }
   }

.nvmrc

@@ -0,0 +1 @@
+24

.prettierignore

@@ -0,0 +1,8 @@
+node_modules/
+coverage/
+pnpm-lock.yaml
+experiments/*/runs/
+experiments/*/fixtures/
+docs/
+.claude/
+*.md

.prettierrc.json

@@ -0,0 +1,6 @@
+{
+  "singleQuote": true,
+  "printWidth": 120,
+  "trailingComma": "all",
+  "arrowParens": "avoid"
+}