docs-data-type-ref
Warn
Audited by Gen Agent Trust Hub on Sep 12, 2026
Risk Level: MEDIUMCOMMAND_EXECUTIONDYNAMIC_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill uses the Bash tool to execute sbt commands such as sbt mdoc, sbt scalafmtAll, and sbt check. While these are standard build tasks, the ability to run arbitrary sbt tasks involves command execution capabilities.
- [DYNAMIC_EXECUTION]: The workflow involves generating Scala code snippets within Markdown files and then executing them using the mdoc tool via sbt. This constitutes a write-then-execute pattern for generated code during the documentation verification phase.
- [INDIRECT_PROMPT_INJECTION]: During the research phase, the agent reads untrusted external data such as source code files, tests, and GitHub history. There is a risk that malicious instructions embedded in these resources could influence the agent's behavior.
- Ingestion points: Step 1 reads source files, tests, and GitHub history using tools like Read, Grep, and Bash.
- Boundary markers: No explicit delimiters are described to separate untrusted source code from the agent's instructions.
- Capability inventory: The skill can execute shell commands via Bash(sbt:*) and write files to the local filesystem.
- Sanitization: No sanitization of the extracted source code or GitHub metadata is described before it is processed by the agent.
Audit Metadata