crit-cli
Pass
Audited by Gen Agent Trust Hub on May 11, 2026
Risk Level: SAFE
Full Analysis
- [COMMAND_EXECUTION]: The skill provides instructions for executing the
critCLI tool. It includes safety guidance to use single quotes for arguments to prevent shell injection and suggests using JSON files for complex data to ensure safe execution. - [DATA_EXFILTRATION]: The
crit sharecommand is documented for uploading files to a remote service. This functionality is an intended part of the tool's feature set for sharing code reviews and is described transparently. - [PROMPT_INJECTION]: The skill involves reading external content from GitHub PRs and review files, which constitutes a surface for indirect prompt injection. Ingestion points include
crit pulland review JSON files; boundary markers are provided by the structured JSON format; capabilities includecritCLI operations; and sanitization is addressed through shell quoting instructions.
Audit Metadata