lead-scoring-model
Pass
Audited by Gen Agent Trust Hub on Oct 25, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process untrusted external data, such as CRM exports, lead files, and fetched web pages. It includes specific defensive instructions (Golden Rule and Workflow Step 1) that explicitly command the AI to treat this content as data only and never as instructions, reducing the risk of indirect prompt injection attacks.
- [COMMAND_EXECUTION]: The skill requires the execution of a local Python script (
scripts/lead_score.py) to perform lead scoring calculations. This script is part of the skill's package and does not perform network operations or access sensitive system files. The execution is a standard method for offloading complex logic from the LLM to a deterministic script.
Audit Metadata