The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes shell commands to identify project configurations and run test suites. Specifically, it executes cat and grep on configuration files (e.g., package.json, requirements.txt) and invokes test runners such as npx vitest and pytest. These are standard operations for a development-focused agent skill.\n- [INDIRECT_PROMPT_INJECTION]: The skill analyzes user-provided source code to generate tests, creating a potential surface for indirect prompt injection where instructions embedded in the code could influence agent behavior. However, this is a known risk for code analysis tools and is mitigated by the scope of the skill's operations. \n
Ingestion points: The agent reads and analyzes source files (functions, classes, dependencies) provided by the user (SKILL.md, Step 1). \n
Boundary markers: There are no explicit markers or instructions provided to the agent to distinguish between code logic and potentially malicious embedded instructions. \n
Capability inventory: The skill has the capability to execute shell commands for environment detection and running test coverage reports (SKILL.md, Step 2 and Step 6). \n
Sanitization: The skill does not implement specific sanitization or filtering of the analyzed source code before processing it. \n- [DYNAMIC_EXECUTION]: The skill generates executable test files based on the analyzed source code and subsequently instructs the agent to execute these tests using tools like vitest or pytest. This is the primary function of the skill and is performed using standard framework templates. \n- [EXTERNAL_DOWNLOADS]: The instructions suggest installing testing frameworks (e.g., via npm or pip) if they are missing from the project. These tools are well-known and reputable software development dependencies.

test-generator