web-research

Pass

Audited by Gen Agent Trust Hub on Mar 17, 2026

Risk Level: SAFE
Full Analysis
  • Indirect Prompt Injection Surface: The skill involves aggregating external content from the web into local files via subagents. This information is subsequently processed by the main agent for synthesis. While essential for research, this process introduces a surface where instructions embedded in external web content could be encountered. Ingestion points: Web findings are saved to findings_[subtopic].md files in SKILL.md. Boundary markers: The current process does not explicitly require the use of delimiters to separate research findings from instruction context. Capability inventory: The skill leverages write_file, task, and read_file for research management. Sanitization: No explicit sanitization or filtering is specified, which is a common pattern for high-level research coordination.
  • Local File System Interaction: The skill organizes its work within a dedicated research directory relative to the current path. While using write_file and read_file for this purpose is standard, it is important to ensure the agent's environment is configured to restrict file operations to the intended project directories.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 17, 2026, 07:11 AM