budget-optimizer
Pass
Audited by Gen Agent Trust Hub on Aug 20, 2026
Risk Level: SAFE
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill exhibits an attack surface for indirect prompt injection as it processes untrusted data from multiple sources. However, its limited capabilities (Markdown generation and local file writing) mitigate this risk.
- Ingestion points: Ingests user-provided campaign goals, total budget, and performance metrics. It also optionally reads data from external connectors such as
~~influencer database,~~social platform analytics, and~~CRM(SKILL.md). - Boundary markers: None identified. Instructions do not currently include delimiters or specific warnings to ignore instructions embedded in the external data.
- Capability inventory: File system write operations are limited to the skill's specific memory directories (
memory/influencer/budget-optimizer/) and a sharedmemory/hot-cache.mdfile for state persistence. The skill does not possess network access or command execution capabilities (SKILL.md). - Sanitization: No specific sanitization or validation logic for the ingested data was identified.
Audit Metadata