portaly-user
Pass
Audited by Gen Agent Trust Hub on May 8, 2026
Risk Level: SAFEPROMPT_INJECTIONDATA_EXFILTRATIONCOMMAND_EXECUTIONCREDENTIALS_UNSAFE
Full Analysis
- [SAFE]: No malicious patterns such as primary prompt injection, obfuscation, or unauthorized data exfiltration were detected. The skill's operations are transparent and require explicit user consent.
- [DATA_EXFILTRATION]: The skill is designed to read local user records and database schemas to synchronize them with the vendor's API at portaly.ai. This data flow is consistent with the skill's stated purpose and targets the vendor's official domain.
- [COMMAND_EXECUTION]: The skill creates and instructs the user to run local scripts for data migration, ensuring the human user retains control over the execution of administrative tasks.
- [CREDENTIALS_UNSAFE]: The skill identifies and manages Portaly API keys via local environment files, following standard and recommended practices for software integrations.
- [PROMPT_INJECTION]: Analysis of the indirect prompt injection surface: 1. Ingestion points: User registration inputs (referral codes) and database records identified in SKILL.md and scripts/sync_user.mjs. 2. Boundary markers: Not explicitly implemented for the data interpolation within the sync helper. 3. Capability inventory: File creation and network POST operations to portaly.ai. 4. Sanitization: Basic type enforcement for identifiers. The risk is assessed as low given the prescriptive nature of the synchronization tasks.
Audit Metadata