The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses local shell commands (e.g., mkdir -p) to establish the required bdd/ directory structure. This is a standard and expected behavior for a scaffolding utility.
[PROMPT_INJECTION]: The skill includes an indirect prompt injection surface as it ingests untrusted data from project files such as prd/changelog.md, prd/specs/, and various README.md files to synthesize context for scenario generation.
Ingestion points: prd/changelog.md, prd/specs/*.md, **/README.md, and source code files.
Boundary markers: The skill does not explicitly use delimiters or 'ignore' instructions when processing this content.
Capability inventory: The skill can write and append to files (feature files, step definitions, and index files) and create directories.
Sanitization: No specific sanitization or filtering of the extracted text is performed before it is used to drive the generation logic.
[EXTERNAL_DOWNLOADS]: The skill templates reference standard, well-known libraries such as github.com/cucumber/godog and @cucumber/cucumber. These are legitimate dependencies for BDD frameworks and do not pose a security risk in this context.

gherkin