backtest

Pass

Audited by Gen Agent Trust Hub on Apr 29, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill utilizes uvx to fetch and execute the tiportfolio package from a public registry at runtime. This is a standard operational pattern for the utility and the package appears to be a legitimate vendor resource.
  • [COMMAND_EXECUTION]: The skill constructs and executes shell commands through the Bash tool to perform simulations. Although the agent is instructed to map user parameters to specific CLI flags, this pattern creates a risk of command injection if the agent fails to sanitize user-provided strings like ticker symbols or rebalancing ratios.
  • [PROMPT_INJECTION]: The skill contains an indirect prompt injection surface where untrusted user input is ingested and used to build executable commands.
  • Ingestion points: Natural language user requests containing tickers, dates, and allocation ratios in SKILL.md.
  • Boundary markers: The instructions lack explicit delimiters or escape sequences to isolate user-provided data from the command template.
  • Capability inventory: The skill has access to the Bash tool for shell command execution.
  • Sanitization: The instruction relies on the agent's extraction logic to identify parameters but does not provide specific rules for escaping shell metacharacters or validating input strings.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 29, 2026, 08:34 PM
Security Audit — agent-trust-hub — backtest