| name | sod-analyzer |
| description | Detect segregation of duties conflicts in user role assignments. Use this skill for role conflict analysis, four-eyes control review, and SOD violation detection. |
Sod Analyzer
Use this skill when the user asks about segregation of duties, SOD analysis, role conflicts, duties separation, or four-eyes controls.
Goal
Detect user role combinations that violate segregation-of-duties expectations and report the conflicts with risk context.
Workflow
- Confirm the user role dataset and whether built-in or custom conflict rules should be used.
- Gather the user-role input and any custom conflict definitions.
- Review README.md if you need examples of supported input shapes or output style.
- Run main.py with the selected inputs and rule settings.
- Summarize the conflicts, affected users, and the highest-risk combinations first.
Inputs
- users (string, required): Path to JSON file mapping usernames to roles.
- conflicts (string, optional): Path to JSON file with custom conflict pairs.
- builtin_conflicts (boolean, optional): Include the built-in conflict library.
Bundled Files
- main.py: executable logic for the skill.
- README.md: detailed usage, supported inputs, and sample outputs.
- sample_input: bundled example user-role and conflict data.
- skill.yaml: existing repo manifest retained for this project's original packaging.
Guidance
- Prefer built-in conflicts for standard reviews and add custom rules when the environment has specialized duties.
- Keep the explanation tied to independence of control and risk, not just role names.
- If the role dataset is incomplete, flag that the analysis may underreport conflicts.
