pushinator
Pass
Audited by Gen Agent Trust Hub on Mar 18, 2026
Risk Level: SAFEPROMPT_INJECTIONDATA_EXFILTRATION
Full Analysis
- [PROMPT_INJECTION]: Indirect prompt injection vulnerability surface detected in notification message handling.
- Ingestion points: User-provided message strings or external data sources (like CI/CD logs or deployment statuses) are interpolated into the
contentfield of the JSON payload inSKILL.md. - Boundary markers: The skill does not provide explicit delimiters or instructions to the agent to disregard potential commands embedded within the message content.
- Capability inventory: The skill utilizes
curlto transmit data to the external Pushinator API (api.pushinator.com). - Sanitization: No sanitization, validation, or escaping logic is implemented for the data processed into the notification body.
- [DATA_EXFILTRATION]: The skill performs outbound network operations to a non-whitelisted external domain (
api.pushinator.com). This is documented as the intended functional behavior for the notification service and does not involve the transmission of local sensitive files.
Audit Metadata