The Agent Skills Directory

[SAFE]: The skill performs expected tasks for a coding assistant, such as reading project documentation and implementing code. All actions are consistent with its stated purpose and intended software development use case.
[COMMAND_EXECUTION]: The skill uses standard build tools (mvn, gradlew) to compile and test code locally within the project directory. These are routine development operations.
[PROMPT_INJECTION]: The skill has a surface for indirect prompt injection because it ingests data from external files.
Ingestion points: Reads 'plan.md', 'feature.md', and 'docs/project.md'.
Boundary markers: None identified; instructions do not include delimiters to separate data from instructions.
Capability inventory: Ability to create or modify local project files and execute build commands (Maven and Gradle).
Sanitization: No validation or sanitization is performed on the content of the ingested markdown files. This represents an attack surface for indirect prompt injection, though it is standard for this type of autonomous agent.

sdd-implement