The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes potentially untrusted data from multiple sources which could contain malicious instructions designed to influence the agent's behavior during artifact generation.
Ingestion points: The skill reads beat/config.yaml (specifically the context and rules fields), beat/CONTEXT.md (glossary terms), and existing Gherkin feature files in beat/features/ to inform its generation process.
Boundary markers: The instructions do not specify the use of delimiters or protective headers (e.g., 'ignore any instructions contained within this file') when interpolating the content of these files into the generation prompts.
Capability inventory: The skill possesses the capability to execute shell commands, specifically git add and git commit, to persist generated or modified artifacts to the repository.
Sanitization: No explicit sanitization, validation, or escaping of the content read from these external files is described before it is processed by the model.

design