Security audit
Security checks for vulnerabilities and agentic risk
The skill does what it says, but its run instructions use a risky shell command template and send lead-search data to an outside service with limited privacy and compliance guidance.
Review generated commands before running them, protect the GMAPS_SCRAPER_API_KEY, and avoid using untrusted or oddly formatted search terms unless the payload is built with a proper JSON encoder instead of raw shell substitution. Only install if you are comfortable sending targeting criteria and lead/contact data to gmapsscraper.io, and make sure any outreach complies with applicable privacy and anti-spam rules.
SKILL.md:75Shell Command Injection Through Unescaped Search Parameters
The skill requires curl and an external API key, indicating intentional data transmission to a third-party service. External transmission is expected for this skill, but it still presents a real security concern because user-provided search criteria and retrieved lead data will leave the local system, and the skill does not clearly communicate those trust boundaries or key-handling expectations.
env:
- GMAPS_SCRAPER_API_KEY
bins:
- curl
primaryEnv: GMAPS_SCRAPER_API_KEY
envVars:
- name: GMAPS_SCRAPER_API_KEY
The skill is explicitly designed to extract business contact details, score them for outreach, and produce outreach-ready lead lists, but it provides no guidance on privacy, consent, data minimization, or compliance with anti-spam and marketing laws. This creates a realistic risk that users will collect and use personal or contact data in ways that violate policy, contractual restrictions, or local regulations.
The skill instructs users to send search keywords, industry targeting, location criteria, and an API credential to a third-party service via curl, but it does not disclose that this information leaves the local environment or may be logged and processed by that provider. Users may unknowingly transmit sensitive market intelligence, customer targeting plans, or regulated data to an external processor.
The request payload sets "lang": "en", which forces a specific language/locale in the skill behavior. The file does not present this as an opt-in choice or explain why English is required for this skill.
No suspicious patterns detected.