uipath-source-control-guardrails
Pass
Audited by Gen Agent Trust Hub on Jul 31, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill serves a defensive purpose by providing a shell script (
block_dangerous_commands.sh) that acts as a gatekeeper for potentially destructive operations. The script uses standard command-line tools likepython3(for JSON parsing) andtrto safely evaluate command inputs against a blocklist of dangerous patterns. - [COMMAND_EXECUTION]: The skill demonstrates safe command handling. The logic is restricted to pattern matching and does not include any arbitrary shell execution, piped remote scripts, or unsanitized input vulnerabilities.
- [PROMPT_INJECTION]: The instructions in
SKILL.mdare focused on establishing safety boundaries and authority policies. There are no attempts to override agent safety filters or bypass platform constraints. - [DATA_EXFILTRATION]: No network operations or sensitive file access patterns were identified. The script operates entirely on local standard input and output.
Audit Metadata