Security audit
Security checks for vulnerabilities and agentic risk
This skill is a straightforward Oxylabs-backed web search and page scraping helper, with expected third-party API use and no evidence of hidden persistence, destructive behavior, or unrelated data access.
Install only if you are comfortable sending search terms, target URLs, geo options, and fetched page content to Oxylabs. Avoid using it for secrets, private/internal URLs, regulated data, or confidential research unless that third-party use is approved.
The skill description does not warn that user search queries, target URLs, and scraped page content are transmitted to Oxylabs, a third-party service. This creates a meaningful privacy and data-governance risk because users or higher-level agents may pass sensitive prompts, internal URLs, or confidential research terms without informed consent.
Code accesses credential files (SSH keys, AWS credentials, etc.). This could indicate credential theft attempts.
export OXYLABS_AI_STUDIO_API_KEY="YOUR_OXYLABS_AI_STUDIO_API_KEY"
### Via .env file
For `~/.openclaw/.env`, add the assignment:
Code accesses credential files (SSH keys, AWS credentials, etc.). This could indicate credential theft attempts.
def _require_api_key() -> str:
key = os.getenv(API_KEY_ENV, "").strip()
if not key:
env_path = Path.home() / ".openclaw" / ".env"
if env_path.is_file():
for line in env_path.read_text(encoding="utf-8").splitlines():
line = line.strip().removeprefix("export ").strip()
Code accesses credential files (SSH keys, AWS credentials, etc.). This could indicate credential theft attempts.
def _require_api_key() -> str:
key = os.getenv(API_KEY_ENV, "").strip()
if not key:
env_path = Path.home() / ".openclaw" / ".env"
if env_path.is_file():
for line in env_path.read_text(encoding="utf-8").splitlines():
line = line.strip().removeprefix("export ").strip()
The skill declares no explicit tool scope or allowed-tools despite requiring environment access, file reads, and network use. That omission weakens least-privilege boundaries and makes it easier for the skill to be invoked with broader capabilities than users may expect, especially in an agent ecosystem where policy metadata guides runtime restrictions.
The activation language is very broad ('whenever the user wants to search the web, look something up, find current information or news, gather sources and links') and can cause over-triggering on common requests. In an agent setting, this can unintentionally route user data or prompts to the external web-search skill when a narrower or local capability would suffice.
This code sends search queries, URLs, geolocation values, and optionally scraped page content to a third-party service via HTTP requests. Although the module docstring states it uses the Oxylabs REST API, there is no runtime confirmation or explicit user-facing warning in the command handlers before transmitting potentially sensitive input.