perses-datasource-manage

Pass

Audited by Gen Agent Trust Hub on Mar 29, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [PROMPT_INJECTION]: The skill presents an attack surface for indirect prompt injection where malicious input could influence command execution.
  • Ingestion points: User-supplied datasource configuration details (URL, name, project) are ingested during the identification phase and processed into resource definitions.
  • Boundary markers: There are no boundary markers or explicit instructions for the agent to ignore malicious commands embedded in the provided metadata.
  • Capability inventory: The skill utilizes the Bash tool to execute percli and curl commands, which can be vulnerable if user input is directly interpolated into shell templates.
  • Sanitization: The skill lacks instructions for input sanitization or escaping, relying on the agent to handle user-provided strings safely when generating shell commands.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 29, 2026, 05:28 PM