rfm-segments
Pass
Audited by Gen Agent Trust Hub on Oct 28, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONDATA_EXFILTRATIONREMOTE_CODE_EXECUTION
Full Analysis
- [SAFE]: The skill provides a structured methodology for RFM segmentation with no evidence of malicious patterns, obfuscation, or persistence mechanisms.
- [INDIRECT_PROMPT_INJECTION]: The skill identifies a potential attack surface by ingesting external data (purchase history, exports, tickets) but proactively mitigates this risk by instructing the agent to treat such content exclusively as data and never as instructions.
- [DATA_EXFILTRATION]: Strong data privacy practices are enforced, including instructions to use the least personal data required for a task, working from aggregates where possible, and strictly forbidding the storage of credentials or personal records in context files.
- [REMOTE_CODE_EXECUTION]: While the skill references
npxcommands for installing additional components, it explicitly commands the agent to inform the user of these commands rather than executing them itself, maintaining a human-in-the-loop requirement. - [SAFE]: The skill references established legal and regulatory bodies (ICO, EUR-Lex) for guidance on data processing and profiling, which are trusted information sources.
Audit Metadata