meta-ads-planning
Pass
Audited by Gen Agent Trust Hub on Jun 23, 2026
Risk Level: SAFE
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and analyze significant amounts of external data, such as Meta Ads Manager exports (CSV/XLSX), CRM reports, and technical diagnostics. This ingestion surface could potentially be used to deliver indirect prompt injections if malicious instructions were embedded in the data (e.g., in campaign names or CRM records). However, the risk is negligible as the skill only generates textual advice and lacks any automated or dangerous execution capabilities.
- Ingestion points: The skill explicitly guides the agent to process exports, tables, and reports from the Meta platform and external sales tools (File: SKILL.md, references/account-data-diagnostics.md).
- Boundary markers: No explicit delimiters or instructions to ignore embedded commands are included in the prompt for processing this data.
- Capability inventory: The agent's capabilities are limited to producing strategic plans, diagnostics, and creative briefs in markdown format.
- Sanitization: No validation or sanitization of input data is specified in the instructions.
Audit Metadata