The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes shell commands such as grep to find specific code patterns and utilizes various language-specific tools including bun, npx, pytest, go, and cargo for linting and test execution.
[EXTERNAL_DOWNLOADS]: It invokes npx eslint, which may download packages from the official NPM registry if they are not already cached in the environment.
[REMOTE_CODE_EXECUTION]: Running the project's test suite via pytest, go test, or cargo test involves the execution of code present in the local repository.
[PROMPT_INJECTION]: This skill possesses an attack surface for indirect prompt injection. It reads project source code to identify and modify defensive programming patterns, which could allow malicious instructions embedded in comments or strings within those files to influence the agent's behavior.
Ingestion points: Reads local project files using grep and eslint (SKILL.md).
Boundary markers: None identified in the provided instructions.
Capability inventory: Filesystem modification, shell command execution, and test suite invocation across multiple environments (SKILL.md).
Sanitization: No explicit sanitization or filtering of the source code content is performed prior to processing by the agent.

cleanup-defensive