The Agent Skills Directory

[SAFE]: The skill does not exhibit any malicious patterns such as credential exfiltration, obfuscation, or unauthorized privilege escalation. Its operations are consistent with its stated purpose as a developer tool.
[COMMAND_EXECUTION]: The skill facilitates the execution of shell scripts and build tools (Gradle, Maven) to manage Kafka infrastructure and run applications. The provided scripts, such as teardown.sh, include robust path validation logic (using realpath) to ensure they do not operate on sensitive system directories like the root or home directory.
[EXTERNAL_DOWNLOADS]: The skill references software dependencies from industry-standard repositories, specifically Maven Central and Confluent's official package registry. These are well-known, trusted sources for Java and Kafka development.
[DATA_EXPOSURE]: The skill manages configuration and credentials through .env files. This is a standard security practice for local development, providing a clear separation between application code and environment-specific secrets, while explicitly instructing the agent not to expose these values in responses.
[INDIRECT_PROMPT_INJECTION]: The skill presents an attack surface for indirect prompt injection because it instructs the agent to read and analyze application logs which may contain data from untrusted external Kafka topics.
Ingestion points: Log reading operations described in references/verification.md and references/debugging.md.
Boundary markers: No specific delimiters or "ignore embedded instructions" warnings are configured for the log analysis phase.
Capability inventory: Execution of shell scripts (create-topics.sh, teardown.sh) and build system commands (gradlew).
Sanitization: There is no explicit sanitization or filtering of log content before it is processed by the agent's context.

kafka-streams-programming