background-researcher-agent

Pass

Audited by Gen Agent Trust Hub on Mar 15, 2026

Risk Level: SAFEPROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
  • [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection due to its core function of ingesting and synthesizing information from untrusted external web sources.
  • Ingestion points: The agent is instructed to read content from technical blogs, GitHub repositories, and Stack Overflow (SKILL.md).
  • Boundary markers: Absent. There are no instructions to use delimiters or to disregard embedded directives within the gathered research data.
  • Capability inventory: The agent has file system access via shell commands and internal RAG search tools.
  • Sanitization: Absent. External findings are directly synthesized into actionable research reports and implementation guides.
  • [COMMAND_EXECUTION]: The skill includes instructions to execute shell commands for codebase analysis which could lead to unintended data exposure.
  • Evidence: The protocol suggests using cat config/*.yaml | head -50 to review configuration patterns. While functional for research, this command could expose hardcoded credentials or sensitive environment details if they are stored within YAML configuration files.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 15, 2026, 11:08 AM