go-testing
Pass
Audited by Gen Agent Trust Hub on Sep 15, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill contains educational material and code snippets following standard Go testing conventions. All examples use the Go standard library and do not introduce external dependencies or insecure practices. The documentation describes idiomatic ways to access environment variables for test configuration (e.g., TEST_DATABASE_URL), which is a standard pattern for integration testing.
- [INDIRECT_PROMPT_INJECTION]: The skill documents the Go fuzz testing framework (testing.F), which by design involves processing untrusted, generated inputs. This represents a documented attack surface for any agent interacting with test logs, but the skill itself provides appropriate safety patterns for this methodology.
- Ingestion points: Fuzz target functions in references/benchmarks-and-fuzzing.md that accept arbitrary string or byte parameters.
- Boundary markers: The Go testing framework provides runtime isolation for tests; no specific prompt boundary markers are required for this code-level testing environment.
- Capability inventory: The skill describes the use of the go test toolchain, which involves command execution and file system access for reports and test data.
- Sanitization: Examples demonstrate the use of t.Skip() to handle and ignore structurally invalid fuzzing inputs.
Audit Metadata