linkedin-ads-creative
Pass
Audited by Gen Agent Trust Hub on Jul 22, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill references resources belonging to the vendor Cold-IQ, including an API endpoint (api.coldiq.com) and an MCP package (@coldiq/mcp). These are standard vendor integrations and do not pose a security risk.
- [INDIRECT_PROMPT_INJECTION]: The skill incorporates external data from LinkedIn via MCP tools, which represents an indirect prompt injection surface. Malicious content in the ingested data (such as competitor ad copy) could potentially influence the agent.
- Ingestion points: Data retrieved via
mcp__coldiq__search_ads,mcp__coldiq__find_people, andmcp__coldiq__enrich_companytools in SKILL.md. - Boundary markers: None specified in the instructions to separate external data from system instructions.
- Capability inventory: The skill focuses on strategic recommendations and format selection; no dangerous command execution is present in the skill itself.
- Sanitization: No filtering or escaping of the external tool output is defined in the instructions.
Audit Metadata