parallel-deep-research

Pass

Audited by Gen Agent Trust Hub on Apr 17, 2026

Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill executes shell commands using parallel-cli and interpolates the user-provided $ARGUMENTS directly into the command string. While the execution is restricted to the parallel-cli toolset via the allowed-tools configuration, unvalidated user input in shell commands remains a potential injection surface.
  • [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it processes and displays research summaries derived from external, untrusted sources.
  • Ingestion points: The executive summary and report files generated by parallel-cli research poll as described in Step 2.
  • Boundary markers: No explicit delimiters or warnings are used to separate research data from the agent's instructions.
  • Capability inventory: The skill has access to shell execution via parallel-cli and file system writes for saving research reports.
  • Sanitization: There is no evidence of sanitization or filtering applied to the external data before it is presented to the user or stored in the context.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 17, 2026, 10:29 AM