The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses the scalafmt CLI tool for project auditing and formatting. This is the intended primary functionality. The skill provides safety-oriented instructions to probe the environment (scalafmt --version) and perform read-only checks (--check .) before executing any modifications.
[INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process Scala source files (.scala, .sc), which creates a standard attack surface for indirect prompt injection common to developer tooling.
Ingestion points: Local project files including .scala, .sc, and build.sbt (SKILL.md, references/tooling/toolchain.md).
Boundary markers: While explicit markdown delimiters for file content are not specified, the skill mandates a strict output template to maintain context separation.
Capability inventory: The skill's actions are restricted to local file reading, formatting via scalafmt, and running existing project build/test commands.
Sanitization: No internal sanitization is performed on file contents, but the skill includes explicit rules against silent modifications and manual override of safety violations.

codeguard-scala