research

Warn

Audited by Gen Agent Trust Hub on Feb 17, 2026

Risk Level: MEDIUMCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • COMMAND_EXECUTION (MEDIUM): The skill utilizes the gemini bash command with a dynamically generated search prompt: gemini -m gemini-2.5-flash -p "...your search prompt...". There is no evidence of sanitization for shell metacharacters (e.g., ;, &, |), which could lead to arbitrary command execution if an attacker-controlled input influences the search prompt.
  • PROMPT_INJECTION (LOW): The skill is vulnerable to Indirect Prompt Injection (Category 8) due to its data processing pipeline.
  • Ingestion points: External content from WebSearch, docs-seeker (GitHub), and the output of the gemini research command.
  • Boundary markers: Absent. There are no specified delimiters or instructions for the agent to ignore instructions embedded within the research data.
  • Capability inventory: File system write access (./plans/ directory) and bash command execution (gemini).
  • Sanitization: Absent. External data is analyzed and synthesized into a markdown report without explicit escaping or validation steps.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Feb 17, 2026, 06:25 PM