data-synthesis
Pass
Audited by Gen Agent Trust Hub on Feb 26, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: No malicious patterns or security vulnerabilities were identified in the skill.
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to process data from external research sources and sub-agents, creating a surface for potential indirect prompt injection. However, the skill lacks the necessary capabilities (such as shell execution or network access) to be exploited. * Ingestion points: Research outputs from sub-agents and files such as 'agent_1_findings.md'. * Boundary markers: Absent in the workflow instructions, though output templates use Markdown headers. * Capability inventory: None. The skill consists only of text instructions and templates for data organization. * Sanitization: None.
- [DATA_EXFILTRATION]: No network operations (curl, wget) or attempts to access sensitive system files (e.g., SSH keys, AWS credentials) were found.
- [COMMAND_EXECUTION]: The skill does not contain shell commands, scripts, or instructions to execute arbitrary code.
- [PROMPT_INJECTION]: The instructions do not contain markers for overriding safety filters or disregarding system prompts.
Audit Metadata