Security audit
Security checks for vulnerabilities and agentic risk
This skill is a straightforward connector for chatting with Pallio AI personas, with expected remote message transmission and no hidden install, persistence, or local data access behavior found.
Before installing, treat Pallio as an external AI service: do not send sensitive, confidential, or regulated information unless you are comfortable with Pallio receiving it. The skill appears limited to widget chat calls and does not install code or request local secrets.
Data is being sent to an external URL. This could be legitimate telemetry or data exfiltration. Manual review is recommended.
Before sending any messages, initialize a session to get an authentication token.
curl -s "https://pallioai.com/api/widget/init/$PALLIO_PERSONA_ID"
The response contains:
The skill explicitly instructs sending the user's message and full conversation history to a third-party Pallio endpoint, but it does not require any user-facing disclosure, consent, or warning about data leaving the local environment. Users may unknowingly transmit sensitive prompts, personal data, or proprietary information to an external service, increasing privacy and compliance risk.
The skill asserts that the persona 'will not fabricate answers,' but the documented implementation is just a remote chat API call with no technical control, validation step, or confidence gating that would enforce grounded-only responses. This can mislead users into overtrusting outputs from an external AI system and may cause unsafe reliance on incorrect information.
No suspicious patterns detected.