testing-essentials
Pass
Audited by Gen Agent Trust Hub on Sep 30, 2026
Risk Level: SAFE
Full Analysis
- [PROMPT_INJECTION]: No attempts to override agent behavior or bypass safety filters were detected. The instructions are focused on providing testing structures and quality gates.
- [DATA_EXPOSURE_AND_EXFILTRATION]: No access to sensitive file paths or unauthorized network operations were found. The use of a dummy password in the fixture pattern (
hello world!) is a standard testing practice and does not expose real credentials. - [OBFUSCATION]: No hidden or encoded content, homoglyphs, or zero-width characters were detected in any of the files.
- [REMOTE_CODE_EXECUTION]: The skill does not download or execute remote scripts. It provides static Elixir code templates and refers to local development tools.
- [COMMAND_EXECUTION]: The skill references standard Elixir build and quality tools (
mix test,mix format,mix credo,mix dialyzer) which are appropriate and expected for a testing and code-quality skill. - [INDIRECT_PROMPT_INJECTION]: The skill provides templates for the agent to follow when generating tests. It does not possess a vulnerability surface for ingesting or processing untrusted external data.
Audit Metadata