The Agent Skills Directory

Command Execution: The skill utilizes standard git commands (git remote get-url origin and git rev-parse --abbrev-ref HEAD) to identify the repository's origin and current branch. This metadata is used to generate accurate remote citations for the user.
Indirect Prompt Injection: The skill's primary function involves reading untrusted content from source code repositories. This presents a potential surface for indirect prompt injection if files contain instructions intended to influence the agent's behavior.
Ingestion points: Source files retrieved during the search and read procedures (SKILL.md).
Boundary markers: No explicit delimiters or instructions are provided to the agent to distinguish executable code or comments from embedded natural language instructions.
Capability inventory: Includes file system reading, local command execution (git), and synthesis of external data into responses.
Sanitization: No specific sanitization, validation, or filtering of file content is described before the information is synthesized into the final answer.

wiki-qa