db-dump-dt

Pass

Audited by Gen Agent Trust Hub on Sep 7, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill loads configuration from a .v8-project.json file in the project workspace, creating a potential vector for indirect parameter injection.
  • Ingestion points: The scripts/db-dump-dt.ps1 and scripts/db-dump-dt.py scripts search the directory hierarchy for .v8-project.json to resolve the v8path and additional CLI arguments.
  • Boundary markers: The skill does not use specific boundary markers for the imported configuration data, but it validates the resulting arguments against a list of prohibited keys.
  • Capability inventory: The skill possesses command execution capabilities through subprocess.run (Python) and System.Diagnostics.Process (PowerShell), which are used to invoke the 1C platform.
  • Sanitization: The scripts include logic to identify and redact sensitive tokens (like -Password) from command-line strings before they are printed to the console.
  • [COMMAND_EXECUTION]: The skill constructs and executes system commands targeting the 1C:Enterprise platform binaries.
  • The command-line construction process incorporates values from user arguments and the workspace configuration file.
  • Validation logic is present to ensure that supplemental arguments do not conflict with or override the skill's intended operation.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 7, 2026, 10:28 PM
Security Audit — agent-trust-hub — db-dump-dt