Suno AI

Security checks across static analysis, malware telemetry, and agentic risk

Overview

This skill is built to automate Suno with saved browser sessions while evading CAPTCHA/automation checks, and it accesses local cookies and API keys beyond what the metadata declares.

Avoid installing this skill unless you fully understand and accept the risks of automated CAPTCHA bypass, saved Suno browser sessions, local cookie extraction, and OpenRouter API usage. If you test it, use a dedicated Suno account and browser profile, remove saved cookies afterward, monitor OpenRouter charges, and do not use it on accounts you cannot risk losing.

Static analysis

No static analysis findings were reported for this release.

VirusTotal

No VirusTotal findings for this skill version.

Malicious
0
Suspicious
0
Harmless
0
Undetected
66
View on VirusTotal

Risk analysis

Artifact-based informational review of SKILL.md, metadata, install specs, static scan signals, and capability signals. ClawScan does not execute the skill or run runtime probes.

What this means

Using this skill may violate Suno or hCaptcha anti-automation rules and could put the user's Suno account, credits, or access at risk.

Why it was flagged

The skill patches an hCaptcha challenger, screenshots CAPTCHA tiles, classifies them with a model, and clicks the selected tiles, showing automated CAPTCHA solving rather than ordinary browser automation.

Skill content
install_hcaptcha_openrouter_patch() ... task_locator.screenshot(type="png", path=tile_path) ... should_click = await _classify_hcaptcha_tile(...) ... await self.click_by_mouse(task_locator)
Recommendation

Do not use automated CAPTCHA solving or anti-bot bypass flows; use an official Suno API or require a human to complete challenges directly.

What this means

The skill may cause automated actions to look like normal human browser activity, increasing account and policy risk.

Why it was flagged

The browser launch intentionally hides automation indicators and weakens sandboxing, which is consistent with anti-detection automation used alongside the CAPTCHA-solving flow.

Skill content
ignore_default_args=["--enable-automation"], args=["--disable-blink-features=AutomationControlled", "--no-sandbox", "--disable-dev-shm-usage"] ... Object.defineProperty(navigator, 'webdriver', { get: () => undefined });
Recommendation

Remove stealth/anti-detection settings and avoid weakening browser sandboxing; make any browser automation transparent and user-controlled.

What this means

A copied Suno session cookie can let automation act as the logged-in user and consume credits or access account information.

Why it was flagged

The skill can read the user's live Chromium cookie database and extract Suno cookies, a high-impact session credential source not declared in the registry metadata.

Skill content
DEFAULT_LIVE_CHROMIUM_COOKIE_DB = Path.home() / "snap" / "chromium" / "common" / "chromium" / "Default" / "Cookies" ... SELECT ... FROM cookies WHERE host_key LIKE '%suno.com%'
Recommendation

Require explicit opt-in before reading browser cookie stores, use a dedicated browser profile, declare the credential/config paths, and disable live browser cookie export by default.

What this means

The skill may spend or expose use of a user's OpenRouter account for CAPTCHA-solving without the credential requirement being clear at install time.

Why it was flagged

The code reads OpenRouter credentials from environment variables or the global OpenClaw .env file, while the skill metadata declares no required environment variables or primary credential.

Skill content
for name in ("OPENROUTER_API_KEY", "OPENCLAW_OPENROUTER_API_KEY") ... OPENCLAW_ENV = Path.home() / ".openclaw" / ".env"
Recommendation

Declare OpenRouter credentials in metadata, ask for user approval before use, and avoid reading global credential files implicitly.

What this means

Images from the browser automation flow leave the local machine for a third-party AI provider, which may have privacy, cost, and account-policy implications.

Why it was flagged

CAPTCHA or page-derived screenshots are converted to image data URLs and sent to OpenRouter, but the SKILL.md does not clearly disclose this external data flow.

Skill content
contents.append({"type": "image_url", "image_url": {"url": _image_to_data_url(image)}}) ... client.post("https://openrouter.ai/api/v1/chat/completions", headers={"Authorization": f"Bearer {self._api_key}"}, json=payload)
Recommendation

Document exactly what data is sent to OpenRouter, require explicit consent, and provide an option to disable third-party image solving.

What this means

The reviewed files do not fully describe which dependency versions will execute on the user's machine.

Why it was flagged

The runnable behavior depends on a local ~/.suno virtual environment that is not installed or pinned by the supplied artifacts.

Skill content
PYTHON_BIN="${SUNO_PYTHON_BIN:-$HOME/.suno/venv/bin/python}" ... exec "$PYTHON_BIN" "$(dirname "$0")/generate_song.py" "$@"
Recommendation

Provide a complete install specification, dependency lockfile, and provenance for the Python runtime and packages.