The Agent Skills Directory

[COMMAND_EXECUTION]: The skill utilizes the Bash tool to perform compilation and test execution tasks (e.g., mvn test-compile, gradle testClasses). These operations are used solely to verify the integrity and correctness of the generated test code, which is a standard functional requirement for this type of tool.
[PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface because it analyzes external source code and dependencies provided by the user.
Ingestion points: Source code files, dependency classes, and existing tests are read into the agent's context during the initial analysis phase (SKILL.md, Step 1).
Boundary markers: The instructions do not define specific delimiters or warnings to ignore instructions found within the analyzed data.
Capability inventory: The agent has the ability to write files to the filesystem and execute shell commands via Bash for build purposes.
Sanitization: No specific sanitization or filtering of the input code is mentioned. However, the risk is mitigated by the human review checkpoint (SKILL.md, Step 3) where the agent must ask for user permission before proceeding from the analysis phase to the generation and execution phases.

generate-tests