bdd
Pass
Audited by Gen Agent Trust Hub on Sep 17, 2026
Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: Provides detailed instructions and guides for installing development tools such as Vitest and Playwright using package managers like npm, pnpm, and yarn. These commands are standard for project setup.
- [EXTERNAL_DOWNLOADS]: References official documentation, GitHub repositories, and package registries for industry-standard testing frameworks (Vitest, Playwright) and their associated BDD libraries. All sources are well-known and reputable within the software development community.
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process user-provided text, such as acceptance criteria and existing feature files, to generate or refactor BDD specifications.
- Ingestion points: Ingests untrusted text describing business requirements or code snippets from the user prompt.
- Boundary markers: No explicit delimiters or warnings to ignore embedded instructions are provided in the skill's instructions.
- Capability inventory: The skill generates markdown (Gherkin) and TypeScript code; it does not have the capability to execute shell commands directly or perform network operations on behalf of the user beyond the provided installation guides.
- Sanitization: The skill relies on the underlying LLM's safety guardrails to process external content without executing malicious instructions contained within.
Audit Metadata