meta-ads-audit
Pass
Audited by Gen Agent Trust Hub on Aug 12, 2026
Risk Level: SAFECOMMAND_EXECUTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill connects to and retrieves data from Meta's official native ads MCP server (mcp.facebook.com). This is a well-known, trusted service used for the skill's primary function.
- [COMMAND_EXECUTION]: The skill is configured to write audit reports to the local filesystem (e.g.,
./meta-ads-audit/audit-<account>.md). This is an expected behavior for generating persistent reports. - [DATA_EXFILTRATION]: While the skill processes advertising data from external APIs, it does not demonstrate any patterns of exfiltrating sensitive credentials or data to unauthorized third-party domains. All communications are directed to official Meta infrastructure.
- [INDIRECT_PROMPT_INJECTION]: The skill ingests data from ad entities (names, campaign titles). There is a theoretical surface for indirect prompt injection if malicious instructions are embedded in ad metadata; however, the skill's logic focuses on quantitative extraction and uses a rigid output schema, which significantly mitigates the risk of the agent following embedded instructions.
Audit Metadata