linkedin-article-writer
Pass
Audited by Gen Agent Trust Hub on Apr 13, 2026
Risk Level: SAFE
Full Analysis
- [PROMPT_INJECTION]: The skill ingests untrusted data from the Tiger Den platform and raw user drafts (SKILL.md). Ingestion points include the
get_content_texttool and direct text input. While explicit boundary markers are absent, the skill mitigates injection risk by instructing the agent to extract useful context from tool-returned content rather than executing it as a direct command. - [DATA_EXFILTRATION]: The skill accesses a 'No Fly List' of customers to ensure compliance with public referencing policies. Although the agent is instructed to inform users when a specific customer is restricted, this is a functional requirement for business safety and does not constitute malicious data exfiltration.
- [EXTERNAL_DOWNLOADS]: The skill retrieves brand voice guidelines and marketing context from the vendor's infrastructure via the Tiger Den MCP server. These references are standard operational dependencies for the skill's intended purpose.
Audit Metadata