multi-source-data-conflation
Pass
Audited by Gen Agent Trust Hub on Sep 24, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process data from multiple external and potentially untrusted sources such as APIs, CRMs, and analytics platforms. This creates a surface for indirect prompt injection where malicious instructions could be embedded in the source data and subsequently processed by an agent.
- Ingestion points: Data enters the system via external API calls (e.g.,
crm_api.get_customers,support_api.get_tickets) and message queues (kafka.subscribe) as described in SKILL.md. - Boundary markers: The provided code templates do not implement delimiters or explicit instructions for the agent to ignore potentially malicious content within the merged data.
- Capability inventory: The skill demonstrates capabilities for database interaction (
master_db.query,master_db.update,master_db.bulk_insert) and network communications (parallel API fetching usingasyncio.gather). - Sanitization: While basic data cleansing and normalization logic is included (e.g., case standardization and phone number formatting), there is no specific sanitization or escaping of content to prevent it from influencing an LLM's downstream behavior.
Audit Metadata