z-go-modernize
Pass
Audited by Gen Agent Trust Hub on Oct 13, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes external Go source code, creating an ingestion point for untrusted data. 1. Ingestion points: Analyzes Go source files, go.mod files, and toolchain directives. 2. Boundary markers: No explicit delimiters are defined to separate code from agent instructions. 3. Capability inventory: Uses shell commands for verification, including go vet, golangci-lint, govulncheck, and go test. 4. Sanitization: None mentioned; the tool chain is applied directly to the source code.
- [COMMAND_EXECUTION]: Includes a verify section with standard toolchain commands: go vet, golangci-lint, govulncheck, and go test, which are appropriate for Go development tasks.
Audit Metadata