The Agent Skills Directory

[SAFE]: The skill is documentation-based and contains no executable scripts, source code, or binary files. It provides a methodology for software verification without introducing programmatic risk.
[SAFE]: No external dependencies (Python or Node.js) are defined, and the skill explicitly states it has no runtime dependencies in the metadata.
[SAFE]: No patterns of obfuscation, command execution, or network exfiltration were identified. Instructions focus on standard engineering practices such as TDD, regression testing, and evidence collection.
[SAFE]: The engineering principles defined in the skill include security-positive guidelines, such as principle 11 which instructs to load only the minimum relevant context and principle 10 which advises preserving unrelated working-tree changes.

accept