The Agent Skills Directory

[COMMAND_EXECUTION]: The skill is designed to run shell commands for various testing frameworks, including pytest, npx jest, mvn test, go test, and cargo test.
[DYNAMIC_EXECUTION]: The skill generates new test scripts based on analyzed source code and executes them using standard runners. This facilitates the execution of dynamically created code.
[INDIRECT_PROMPT_INJECTION]: The skill has an attack surface for indirect prompt injection because it processes untrusted source code and then executes tests based on that analysis.
Ingestion points: The skill reads local source files, directory structures, and existing test files (SKILL.md).
Boundary markers: Missing explicit delimiters or instructions to ignore potential instructions embedded in the code it analyzes.
Capability inventory: The skill utilizes shell execution for running tests and coverage instrumentation across multiple programming languages (SKILL.md).
Sanitization: There is no mention of sanitizing or validating the input files to prevent malicious code or instructions from influencing the agent's actions.

testing