The Agent Skills Directory

[SAFE]: The skill body and referenced files contain purely educational content regarding software testing methodologies and design patterns.
[INDIRECT_PROMPT_INJECTION]: The skill suggests that the agent read a local CONTEXT.md file to align its vocabulary with the project's domain language. This introduces a potential surface for indirect prompt injection if the local repository contains untrusted content.
Ingestion points: Instructions to read CONTEXT.md from the workspace.
Boundary markers: Not present.
Capability inventory: The skill provides guidance for agents performing coding tasks, which usually include file modification and command execution capabilities.
Sanitization: None.
[COMMAND_EXECUTION]: While the skill discusses testing and code implementation, it does not include any automated shell command execution (!command) or script invocation patterns.

tdd