@houwert/conductor 0.5.0 → 0.6.0

package/dist/index.js

@@ -32,6 +32,7 @@ const run_parallel_js_1 = require("./commands/run-parallel.js");
 const foreground_app_js_1 = require("./commands/foreground-app.js");
 const list_apps_js_1 = require("./commands/list-apps.js");
 const copy_app_js_1 = require("./commands/copy-app.js");
+const download_app_js_1 = require("./commands/download-app.js");
 const install_app_js_1 = require("./commands/install-app.js");
 const erase_text_js_1 = require("./commands/erase-text.js");
 const assert_not_visible_js_1 = require("./commands/assert-not-visible.js");
@@ -41,21 +42,25 @@ const scroll_until_visible_js_1 = require("./commands/scroll-until-visible.js");
 const set_location_js_1 = require("./commands/set-location.js");
 const set_orientation_js_1 = require("./commands/set-orientation.js");
 const start_device_js_1 = require("./commands/start-device.js");
+const delete_device_js_1 = require("./commands/delete-device.js");
+const logs_js_1 = require("./commands/logs.js");
 const device_picker_js_1 = require("./device-picker.js");
 const update_check_js_1 = require("./update-check.js");
 const COMMAND_HELP = {
     'start-device': start_device_js_1.HELP,
+    'delete-device': delete_device_js_1.HELP,
     'list-devices': list_devices_js_1.HELP,
     'foreground-app': foreground_app_js_1.HELP,
     'list-apps': list_apps_js_1.HELP,
     'copy-app': copy_app_js_1.HELP,
+    'download-app': download_app_js_1.HELP,
     'install-app': install_app_js_1.HELP,
     'launch-app': launch_app_js_1.HELP,
     'stop-app': stop_app_js_1.HELP,
     'clear-state': clear_state_js_1.HELP,
     'uninstall-app': uninstall_app_js_1.HELP,
-    tap: tap_js_1.HELP,
-    type: type_js_1.HELP,
+    'tap-on': tap_js_1.HELP,
+    'input-text': type_js_1.HELP,
     'erase-text': erase_text_js_1.HELP,
     back: back_js_1.HELP,
     'press-key': press_key_js_1.HELP,
@@ -68,23 +73,27 @@ const COMMAND_HELP = {
     'open-link': open_link_js_1.HELP,
     'set-location': set_location_js_1.HELP,
     'set-orientation': set_orientation_js_1.HELP,
-    screenshot: screenshot_js_1.HELP,
+    'take-screenshot': screenshot_js_1.HELP,
     inspect: inspect_js_1.HELP,
     focused: focused_js_1.HELP,
     'run-flow': run_flow_js_1.HELP,
     'run-flow-inline': run_flow_inline_js_1.HELP,
     session: session_js_1.HELP,
     'cheat-sheet': cheat_sheet_js_1.HELP,
-    install: install_js_1.HELP,
+    'install-plugin': install_js_1.HELP_INSTALL_PLUGIN,
+    'install-skills': install_js_1.HELP_INSTALL_SKILLS,
+    'install-web': install_js_1.HELP_INSTALL_WEB,
     'daemon-start': daemon_js_1.HELP_DAEMON_START,
     'daemon-stop': daemon_js_1.HELP_DAEMON_STOP,
     'daemon-status': daemon_js_1.HELP_DAEMON_STATUS,
     'device-pool': device_pool_js_1.HELP,
     'run-parallel': run_parallel_js_1.HELP,
+    logs: logs_js_1.HELP,
 };
 const OPTIONS_HELP = `Options:
   --device <id>     Target device ID (also keys the session and daemon)
   --device-name <n> Target a booted device by name (resolved to ID from booted devices)
+  --platform <p>    Filter to devices of this platform (ios, android, tvos, web)
   --json            Output as machine-readable JSON
   --verbose, -v     Log daemon calls, fallbacks, and raw output
   --help, -h        Show this help`;
@@ -103,7 +112,6 @@ async function main() {
             'clear',
             'list',
             'verbose',
-            'skills',
             'check',
             'all',
             'acquire',
@@ -140,8 +148,12 @@ async function main() {
             'name',
             'device-name',
             'device-type',
+            'browser',
             'from',
             'to',
+            'source',
+            'metro-port',
+            'level',
         ],
         alias: { h: 'help', v: 'verbose' },
     });
@@ -164,11 +176,18 @@ async function main() {
     const NO_DEVICE_COMMANDS = new Set([
         'list-devices',
         'start-device',
+        'delete-device',
         'cheat-sheet',
-        'install',
+        'install-plugin',
+        'install-skills',
+        'install-web',
         'copy-app',
         'device-pool',
         'run-parallel',
+        // `logs --list` and `logs --source metro` only query Metro on localhost — no device needed
+        ...(command === 'logs' && (argv['list'] || argv['source'] === 'metro') ? ['logs'] : []),
+        // `daemon-stop --all` stops every daemon — no device needed
+        ...(command === 'daemon-stop' && argv['all'] ? ['daemon-stop'] : []),
     ]);
     if (!NO_DEVICE_COMMANDS.has(command) && !COMMAND_HELP[command]) {
         console.error(`Unknown command: ${command}`);
@@ -196,7 +215,8 @@ async function main() {
             sessionName = match.id;
         }
         else {
-            sessionName = explicitDevice ?? (await (0, device_picker_js_1.pickDevice)()) ?? 'default';
+            sessionName =
+                explicitDevice ?? (await (0, device_picker_js_1.pickDevice)(argv['platform'])) ?? 'default';
         }
     }
     let exitCode = 0;
@@ -207,11 +227,18 @@ async function main() {
                 avd: argv['avd'],
                 name: argv['name'],
                 deviceType: argv['device-type'],
+                browser: argv['browser'],
             });
             break;
         case 'list-devices':
             exitCode = await (0, list_devices_js_1.listDevices)(opts);
             break;
+        case 'delete-device':
+            exitCode = await (0, delete_device_js_1.deleteDevice)(rest[0], opts, {
+                platform: argv['platform'],
+                all: argv['all'],
+            });
+            break;
         case 'foreground-app':
             exitCode = await (0, foreground_app_js_1.foregroundApp)(opts, sessionName);
             break;
@@ -225,6 +252,11 @@ async function main() {
             exitCode = await (0, copy_app_js_1.copyApp)(bundleId, from ?? '', to ?? '', opts);
             break;
         }
+        case 'download-app': {
+            const appId = rest[0] ?? '';
+            exitCode = await (0, download_app_js_1.downloadApp)(appId, argv['output'], opts, sessionName);
+            break;
+        }
         case 'install-app': {
             const appPath = rest[0] ?? '';
             exitCode = await (0, install_app_js_1.installApp)(appPath, opts, sessionName);
@@ -261,7 +293,7 @@ async function main() {
             exitCode = await (0, uninstall_app_js_1.uninstallApp)(appId, opts, sessionName);
             break;
         }
-        case 'tap': {
+        case 'tap-on': {
             const element = rest.join(' ');
             exitCode = await (0, tap_js_1.tap)(element, opts, sessionName, {
                 id: argv['id'],
@@ -281,7 +313,7 @@ async function main() {
             });
             break;
         }
-        case 'type': {
+        case 'input-text': {
             const text = rest.join(' ');
             exitCode = await (0, type_js_1.typeText)(text, opts, sessionName);
             break;
@@ -397,7 +429,7 @@ async function main() {
             exitCode = await (0, set_orientation_js_1.setOrientation)(orientation, opts, sessionName);
             break;
         }
-        case 'screenshot': {
+        case 'take-screenshot': {
             const outPath = argv['output'];
             exitCode = await (0, screenshot_js_1.screenshot)(outPath, opts, sessionName);
             break;
@@ -411,6 +443,18 @@ async function main() {
                 interval: argv['interval'] !== undefined ? Number(argv['interval']) : undefined,
             });
             break;
+        case 'logs':
+            exitCode = await (0, logs_js_1.logs)(opts, sessionName, {
+                source: argv['source'],
+                level: argv['level'],
+                metro: argv['metro'] || argv['metro-port'] !== undefined,
+                metroPort: argv['metro-port'] !== undefined ? Number(argv['metro-port']) : undefined,
+                target: argv['target'] !== undefined ? Number(argv['target']) : undefined,
+                list: argv['list'],
+                recent: argv['recent'] !== undefined ? Number(argv['recent']) : undefined,
+                duration: argv['duration'] !== undefined ? Number(argv['duration']) : undefined,
+            });
+            break;
         case 'run-flow': {
             const file = rest[0] ?? '';
             const rawEnv = argv['env'];
@@ -430,8 +474,14 @@ async function main() {
         case 'cheat-sheet':
             exitCode = await (0, cheat_sheet_js_1.cheatSheet)();
             break;
-        case 'install':
-            exitCode = await (0, install_js_1.installSkills)(opts, argv['skills'], argv['check']);
+        case 'install-plugin':
+            exitCode = await (0, install_js_1.installPluginCli)(opts, argv['check']);
+            break;
+        case 'install-skills':
+            exitCode = await (0, install_js_1.installSkillsCli)(opts, argv['check']);
+            break;
+        case 'install-web':
+            exitCode = await (0, install_js_1.installWebCli)(opts, argv['check'], rest[0]);
             break;
         case 'daemon-start':
             exitCode = await (0, daemon_js_1.daemonStart)(opts, sessionName);

package/dist/runner.js

@@ -11,6 +11,7 @@ const flow_runner_js_1 = require("./drivers/flow-runner.js");
 const verbose_js_1 = require("./verbose.js");
 const ios_js_1 = require("./drivers/ios.js");
 const android_js_1 = require("./drivers/android.js");
+const web_js_1 = require("./drivers/web.js");
 const bootstrap_js_1 = require("./drivers/bootstrap.js");
 const client_js_1 = require("./daemon/client.js");
 /**
@@ -92,8 +93,10 @@ async function getDriver(sessionName = 'default') {
         throw new Error('No device found. Connect a device or start a simulator, then run again.');
     }
     const platform = await (0, bootstrap_js_1.detectPlatform)(deviceId);
-    const port = await (0, bootstrap_js_1.getDriverPort)(platform, deviceId);
-    (0, verbose_js_1.log)(`getDriver: platform=${platform} deviceId=${deviceId} port=${port}`);
+    // Web daemons resolve their own port after generating a unique session ID,
+    // so we defer getDriverPort for them to avoid allocating an unused port.
+    const port = platform !== 'web' ? await (0, bootstrap_js_1.getDriverPort)(platform, deviceId) : 0;
+    (0, verbose_js_1.log)(`getDriver: platform=${platform} deviceId=${deviceId} port=${port || '(deferred)'}`);
     let driver;
     if (platform === 'ios') {
         if (!(await (0, bootstrap_js_1.isPortOpen)(port))) {
@@ -121,6 +124,33 @@ async function getDriver(sessionName = 'default') {
         }
         driver = tvosDriver;
     }
+    else if (platform === 'web') {
+        const browser = (0, bootstrap_js_1.webBrowserName)(deviceId);
+        let webSession = deviceId;
+        if ((0, bootstrap_js_1.isUnqualifiedWebId)(deviceId)) {
+            const existing = await (0, client_js_1.findRunningWebSession)(browser);
+            if (existing) {
+                webSession = existing;
+                (0, verbose_js_1.log)(`Reusing running web session ${webSession}`);
+            }
+            else {
+                webSession = (0, bootstrap_js_1.generateWebSessionId)(browser);
+                (0, verbose_js_1.log)(`Generated new web session ${webSession}`);
+            }
+        }
+        const webPort = await (0, bootstrap_js_1.getDriverPort)('web', webSession);
+        if (!(await (0, bootstrap_js_1.isPortOpen)(webPort))) {
+            (0, verbose_js_1.log)(`Web driver not running — starting daemon for ${webSession}...`);
+            await (0, client_js_1.startDaemon)(webSession);
+            await waitForPort(webPort);
+        }
+        const webDriver = new web_js_1.WebDriver(webPort, '127.0.0.1', webSession);
+        if (!(await webDriver.isAlive())) {
+            throw new Error(`Web browser driver on port ${webPort} is not responding.\n` +
+                `Run: conductor daemon-start --device ${webSession}`);
+        }
+        driver = webDriver;
+    }
     else {
         // Ensure the daemon is running — it handles APK install and driver startup.
         await (0, client_js_1.startDaemon)(deviceId);

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@houwert/conductor",
-  "version": "0.5.0",
+  "version": "0.6.0",
   "description": "CLI tool for mobile app interactions — optimized for AI agents",
   "license": "MIT",
   "repository": {
@@ -35,12 +35,15 @@
     "@grpc/grpc-js": "^1.12.0",
     "@grpc/proto-loader": "^0.7.15",
     "js-yaml": "^4.1.1",
-    "minimist": "^1.2.8"
+    "minimist": "^1.2.8",
+    "playwright-core": "^1.52.0",
+    "ws": "^8.20.0"
   },
   "devDependencies": {
     "@types/js-yaml": "^4.0.9",
     "@types/minimist": "^1.2.5",
     "@types/node": "^20.0.0",
+    "@types/ws": "^8.18.1",
     "@typescript-eslint/eslint-plugin": "^8.56.1",
     "@typescript-eslint/parser": "^8.56.1",
     "eslint": "^10.0.2",

package/skills/conductor/SKILL.md

@@ -1,6 +1,6 @@
 ---
 name: conductor
-version: 0.2.0
+version: 0.6.0
 description: "Token-efficient CLI for mobile UI testing (iOS simulator + Android emulator), designed for AI agents"
 metadata.openclaw:
   category: service
@@ -157,21 +157,21 @@ conductor uninstall-app com.example.myapp --device emulator-5554
 
 ---
 
-### `tap <element>`
+### `tap-on <element>`
 
 Tap a UI element by its text label or accessibility ID. **Try the most obvious text label or ID first** — e.g. `"Sign In"`, `"Submit"`, `"btn_login"`. Only run `inspect` to look up the exact identifier if your first attempt fails. **For icon-only buttons (no visible text label), always run `inspect` first** to find the accessibility ID before tapping.
 
 ```bash
-conductor tap "Sign In"
-conductor tap --id "btn_login"             # match by accessibility ID instead of text
-conductor tap --text "Edit"                # match by text only (not id)
-conductor tap "Next" --index 1             # pick the 2nd match (0-based)
-conductor tap "Add to cart" --long-press
-conductor tap "Like" --double-tap
-conductor tap "Delete" --optional          # do not fail if not found
-conductor tap "Edit" --below "Username"    # tap "Edit" that is below the "Username" element
-conductor tap "Submit" --above "Footer"
-conductor tap ">" --right-of "Email"
+conductor tap-on "Sign In"
+conductor tap-on --id "btn_login"             # match by accessibility ID instead of text
+conductor tap-on --text "Edit"                # match by text only (not id)
+conductor tap-on "Next" --index 1             # pick the 2nd match (0-based)
+conductor tap-on "Add to cart" --long-press
+conductor tap-on "Like" --double-tap
+conductor tap-on "Delete" --optional          # do not fail if not found
+conductor tap-on "Edit" --below "Username"    # tap "Edit" that is below the "Username" element
+conductor tap-on "Submit" --above "Footer"
+conductor tap-on ">" --right-of "Email"
 ```
 
 Flags:
@@ -192,13 +192,13 @@ Flags:
 
 ---
 
-### `type <text>`
+### `input-text <text>`
 
 Type text into the currently focused input field.
 
 ```bash
-conductor type "hello@example.com"
-conductor type "my password"
+conductor input-text "hello@example.com"
+conductor input-text "my password"
 ```
 
 ---
@@ -216,7 +216,7 @@ conductor erase-text 10     # erase 10 characters
 
 ### `back`
 
-Press the Android back button. **Android only** — iOS has no back button. On iOS, always run `inspect` first to find the exact label or accessibility ID of the in-app back button, then use `tap` to press it.
+Press the Android back button. **Android only** — iOS has no back button. On iOS, always run `inspect` first to find the exact label or accessibility ID of the in-app back button, then use `tap-on` to press it.
 
 ```bash
 conductor back
@@ -386,14 +386,14 @@ conductor set-orientation landscape
 
 ---
 
-### `screenshot`
+### `take-screenshot`
 
-Take a screenshot of the current screen. **Prefer `inspect` over `screenshot`** when you need to understand what is on screen — the view hierarchy tells you element text, IDs, and structure without consuming a vision token. Use `screenshot` only for visual evidence, debugging rendering issues, or when a human needs to see the screen.
+Take a screenshot of the current screen. **Prefer `inspect` over `take-screenshot`** when you need to understand what is on screen — the view hierarchy tells you element text, IDs, and structure without consuming a vision token. Use `take-screenshot` only for visual evidence, debugging rendering issues, or when a human needs to see the screen.
 
 ```bash
-conductor screenshot
-conductor screenshot --output /tmp/screen.png
-conductor screenshot --output ./screenshots/login.png
+conductor take-screenshot
+conductor take-screenshot --output /tmp/screen.png
+conductor take-screenshot --output ./screenshots/login.png
 ```
 
 Default output: `./screenshot-<timestamp>.png`
@@ -404,9 +404,9 @@ Default output: `./screenshot-<timestamp>.png`
 
 ### `inspect`
 
-Print the UI element hierarchy of the current screen. Use this when a `tap` or `assert-visible` fails and you need to discover the exact element text or accessibility ID.
+Print the UI element hierarchy of the current screen. Use this when a `tap-on` or `assert-visible` fails and you need to discover the exact element text or accessibility ID.
 
-The hierarchy shows each element's type, text, and accessibility ID (resourceId / accessibilityIdentifier). Use the `text` value or `id` value as the argument to `tap` / `assert-visible`.
+The hierarchy shows each element's type, text, and accessibility ID (resourceId / accessibilityIdentifier). Use the `text` value or `id` value as the argument to `tap-on` / `assert-visible`.
 
 ```bash
 conductor inspect
@@ -414,6 +414,37 @@ conductor inspect
 
 ---
 
+### `logs`
+
+Stream app console/log output in real time. Auto-detects the best log source based on the platform:
+
+- **Web** — captures Playwright browser console events
+- **React Native (iOS/Android)** — connects to the Metro dev server via Chrome DevTools Protocol
+- **Native iOS** — streams from the simulator via `simctl log stream`
+- **Native Android** — streams from the device via `adb logcat`
+
+```bash
+conductor logs                           # auto-detect source and stream
+conductor logs --source metro            # force Metro CDP (for React Native JS logs)
+conductor logs --source device           # force platform-native logs (simctl/logcat)
+conductor logs --level warn              # only show warnings and errors
+conductor logs --level error             # only show errors
+conductor logs --metro-port 8082         # custom Metro dev server port (default: 8081)
+conductor logs --target 1                # connect to a specific Metro debugger target (0-based)
+conductor logs --list                    # list available Metro debugger targets and exit
+conductor logs --list --json             # list targets as JSON (includes appId, logicalDeviceId)
+conductor logs --json                    # NDJSON output (one JSON object per line)
+```
+
+For iOS/Android, auto-detection tries Metro first (2s timeout), then falls back to device logs. When multiple devices share one Metro instance, use `--list` to see available targets, then `--target <n>` to connect to a specific one. `--list` and `--source metro` don't require a running device session.
+
+Exits on Ctrl+C. Each line in `--json` mode is a complete JSON object:
+```json
+{"timestamp":"...","level":"log","message":"User tapped sign-in","stackTrace":null,"source":"metro"}
+```
+
+---
+
 ### `run-flow <file>`
 
 Execute a Maestro YAML flow file.
@@ -460,12 +491,12 @@ conductor cheat-sheet
 
 ---
 
-### `install`
+### `install-skills`
 
 Copy the conductor skill files into the current project's `.claude/skills/conductor/` directory, so AI agents in that project can use the `cheat-sheet` command locally.
 
 ```bash
-conductor install --skills
+conductor install-skills
 ```
 
 ---
@@ -541,13 +572,13 @@ Exits with code 0 if all flows pass, 1 if any fail.
 
 ## Typical Agent Workflow
 
-**Prefer individual CLI commands over `run-flow` / `run-flow-inline`.** Use `tap`, `type`, `scroll`, `swipe`, `assert-visible`, etc. directly — they are faster, give immediate feedback per step, and make failures easier to diagnose. Only reach for `run-flow` or `run-flow-inline` when you need Maestro-specific YAML features (conditional logic, `runScript`, retryTapIfNoChange`, etc.) that have no CLI equivalent.
+**Prefer individual CLI commands over `run-flow` / `run-flow-inline`.** Use `tap-on`, `input-text`, `scroll`, `swipe`, `assert-visible`, etc. directly — they are faster, give immediate feedback per step, and make failures easier to diagnose. Only reach for `run-flow` or `run-flow-inline` when you need Maestro-specific YAML features (conditional logic, `runScript`, retryTapIfNoChange`, etc.) that have no CLI equivalent.
 
-**Always `launch-app` before interacting.** Never attempt to `tap`, `type`, `scroll`, or navigate before the app is launched and the session is set. `launch-app` both opens the app and saves the `appId`/`deviceId` to the session so all subsequent commands know which device and app to target.
+**Always `launch-app` before interacting.** Never attempt to `tap-on`, `input-text`, `scroll`, or navigate before the app is launched and the session is set. `launch-app` both opens the app and saves the `appId`/`deviceId` to the session so all subsequent commands know which device and app to target.
 
-**To understand what is on screen, run `inspect` first** — it gives you element text, IDs, and structure. Only take a `screenshot` when you need visual evidence or are debugging a rendering issue.
+**To understand what is on screen, run `inspect` first** — it gives you element text, IDs, and structure. Only take a `take-screenshot` when you need visual evidence or are debugging a rendering issue.
 
-**Try obvious text labels and IDs first. Only run `inspect` when a `tap` or `assert-visible` fails and you need to discover the exact identifier.**
+**Try obvious text labels and IDs first. Only run `inspect` when a `tap-on` or `assert-visible` fails and you need to discover the exact identifier.**
 
 Single-agent (default session, no flag needed):
 ```bash
@@ -558,17 +589,17 @@ conductor list-devices
 conductor launch-app com.example.myapp --device emulator-5554
 
 # 3. Interact using the most likely text labels or IDs
-conductor tap "Sign In"           # try the obvious label first
-conductor tap "username_field"    # or a guessed test ID
-conductor type "user@example.com"
-conductor tap "password_field"
-conductor type "secret123"
-conductor tap "Login"
+conductor tap-on "Sign In"           # try the obvious label first
+conductor tap-on "username_field"    # or a guessed test ID
+conductor input-text "user@example.com"
+conductor tap-on "password_field"
+conductor input-text "secret123"
+conductor tap-on "Login"
 
 # → If a tap fails, run inspect to find the real identifier:
 conductor inspect
 # → hierarchy shows text="Log in", id="btn_login" — retry with correct value
-conductor tap "Log in"
+conductor tap-on "Log in"
 
 # 4. Assert — try expected text directly
 conductor assert-visible "Dashboard"
@@ -577,7 +608,7 @@ conductor assert-visible "Dashboard"
 conductor inspect
 
 # 6. Screenshot only for visual evidence or rendering checks
-conductor screenshot --output /tmp/tester-ios-iphone-16/post-login.png
+conductor take-screenshot --output /tmp/tester-ios-iphone-16/post-login.png
 
 # 7. Run a full flow only when CLI commands aren't sufficient
 #    Write the YAML to /tmp/<agent-name>/ first — never pass YAML inline, it doesn't work
@@ -674,13 +705,13 @@ conductor launch-app com.example.myapp --device emulator-5554
 **Every subsequent command carries `--device`:**
 ```bash
 # Agent tester-ios-iphone-16
-conductor tap "Sign In"     --device C59D3241-FB6A-4E3B-AE7B-A82D3C933889
-conductor type "user@a.com" --device C59D3241-FB6A-4E3B-AE7B-A82D3C933889
+conductor tap-on "Sign In"     --device C59D3241-FB6A-4E3B-AE7B-A82D3C933889
+conductor input-text "user@a.com" --device C59D3241-FB6A-4E3B-AE7B-A82D3C933889
 conductor inspect           --device C59D3241-FB6A-4E3B-AE7B-A82D3C933889
 
 # Agent tester-android-pixel-6 (runs fully in parallel)
-conductor tap "Sign In"     --device emulator-5554
-conductor type "user@b.com" --device emulator-5554
+conductor tap-on "Sign In"     --device emulator-5554
+conductor input-text "user@b.com" --device emulator-5554
 conductor inspect           --device emulator-5554
 ```
 

package/skills/skills.yaml

@@ -3,6 +3,6 @@ skills:
     path: conductor/SKILL.md
     description: "Token-efficient CLI for mobile UI testing, designed for AI agents"
     category: service
-    version: 0.2.0
+    version: 0.6.0
     requires:
       bins: [conductor]