Hi team, thanks for open-sourcing this skill!
Currently all 20 files in skills/security-audit/ are flat in the root directory. Following common Agent Skill conventions (e.g. Anthropic's skill spec), it would be much cleaner to group auxiliary files into standard subdirectories:
skills/security-audit/
├── SKILL.md
├── references/ # 14 companion & methodology docs (*.md)
├── resources/ # report-schema.json
└── scripts/ # validator scripts & tests (*.cjs)
Main benefits:
- Better context loading: Keeps
SKILL.md as the lightweight entry point so agents/runtimes only pull in domain companions on-demand rather than ingesting all docs upfront.
- Tooling compatibility: Plays nicely with skill loaders and platforms that expect executable code in
scripts/ and reference material in references/.
Happy to open a PR for this with updated relative paths and passing tests if you're open to the change.
Hi team, thanks for open-sourcing this skill!
Currently all 20 files in
skills/security-audit/are flat in the root directory. Following common Agent Skill conventions (e.g. Anthropic's skill spec), it would be much cleaner to group auxiliary files into standard subdirectories:Main benefits:
SKILL.mdas the lightweight entry point so agents/runtimes only pull in domain companions on-demand rather than ingesting all docs upfront.scripts/and reference material inreferences/.Happy to open a PR for this with updated relative paths and passing tests if you're open to the change.