retention-orchestrator

Pass

Audited by Gen Agent Trust Hub on Jun 23, 2026

Risk Level: SAFE
Full Analysis
  • [COMMAND_EXECUTION]: The skill executes local Python scripts to process cohort data and apply decision matrices. These scripts (classify_reason.py and apply_matrix.py) are located within the skill's internal paths and do not involve remote code downloads or elevated privileges.
  • [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data in the form of customer cancellation reasons from Zoho Billing reports. This represents a potential injection surface if customer-provided text contains malicious instructions. However, the risk is mitigated because the skill is strictly 'propose-only,' outputting recommendations to a Markdown file for human review rather than executing autonomous actions or sending data to external communication channels.
  • [DATA_EXFILTRATION]: The skill accesses sensitive subscription and organization data via the Zoho Billing connector. The analysis confirms that this data is processed locally to generate reports and is not sent to any external or non-whitelisted domains. The operations are consistent with the vendor's intended functionality.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 23, 2026, 10:59 AM
Security Audit — agent-trust-hub — retention-orchestrator