linkedin-abm-audit
Pass
Audited by Gen Agent Trust Hub on Jul 28, 2026
Risk Level: SAFEPROMPT_INJECTIONEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONDATA_EXFILTRATION
Full Analysis
- [PROMPT_INJECTION]: Indirect Prompt Injection Surface:
- Ingestion points: Data is ingested from external LinkedIn and CRM sources (e.g., campaign names, ad content, deal names) via the ZenABM connector tools.
- Boundary markers: The instructions lack directives for the agent to use delimiters or protective framing when incorporating external data into reports.
- Capability inventory: The skill has file-writing capabilities (HTML/PDF generation) and is authorized to execute several data retrieval tools.
- Sanitization: No instructions are provided for the agent to sanitize or escape user-controlled data before it is included in the final audit document.
- [EXTERNAL_DOWNLOADS]: Mentions fetching report templates and assets from the vendor's public repository at "github.com/ZENABM/linkedin-abm-skills" to style the output reports.
- [COMMAND_EXECUTION]: Instructions involve using rendering utilities such as "WeasyPrint" or the host system's PDF generation features to convert HTML reports into a downloadable format.
- [DATA_EXFILTRATION]: Handles and summarizes sensitive advertising and sales data, including spend figures, metrics, and pipeline values. This data is processed as part of the core auditing functionality to create the user's report.
Audit Metadata