The Agent Skills Directory

[COMMAND_EXECUTION]: The skill instructs the agent to execute Maven commands such as ./mvnw compile and ./mvnw clean verify. This facilitates the execution of the project's Maven wrapper script, which is a standard development workflow but involves executing shell scripts from the project root.
[INDIRECT_PROMPT_INJECTION]: The skill possesses an attack surface for indirect prompt injection because it is designed to ingest and refactor untrusted source code and build configurations from a local project environment.
Ingestion points: The agent is instructed to read and inspect current project setup, including Java resource classes and Maven build files in SKILL.md and references/402-frameworks-quarkus-rest.md.
Boundary markers: The instructions lack explicit boundary markers or directions for the agent to ignore instructions that might be embedded within comments or strings in the ingested source code.
Capability inventory: The skill combined with the agent's environment provides capabilities to write to the file system and execute shell commands (./mvnw).
Sanitization: No explicit sanitization, validation, or filtering is prescribed for the data ingested from the project files before the agent processes them or triggers a build.

402-frameworks-quarkus-rest