lead-routing
Pass
Audited by Gen Agent Trust Hub on Sep 13, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes inbound lead data from untrusted external sources, which is a potential surface for indirect prompt injection.
- Ingestion points:
SKILL.md(Interview section) andreferences/test-pack-and-rollout.md(The test pack table) specify that lead data enters the workflow via web forms, APIs, and list imports. - Boundary markers: The instructions do not define explicit delimiters or "ignore embedded instructions" warnings to separate untrusted lead data from the agent's operating context.
- Capability inventory: The skill is instructional and defines design logic; it does not contain code for subprocess execution, network operations, or file system writes.
- Sanitization: The skill's 'Failure modes and fixes' section in
SKILL.mdrecommends implementing field validation and enrichment before routing decisions are made.
Audit Metadata