| name | iam-access-review |
| description | Analyse IAM policies for excessive permissions, wildcards, admin-equivalent actions, missing conditions, and inline policies. Use this skill for access review and least-privilege analysis. |
Access Review
Use this skill when the user asks for IAM review, access review, policy review, excessive permissions analysis, wildcard permission review, or least-privilege checks.
Goal
Analyse IAM policy data to identify dangerous permissions, wildcards, admin-equivalent actions, and missing security conditions.
Workflow
- Confirm whether the user is supplying local policy data or wants a cloud-backed mode.
- Gather the policy input and any mode selection.
- Review README.md if you need exact input formats, supported modes, or sample outputs.
- Run main.py with the relevant input, mode, and optional dry-run settings.
- Summarize the most dangerous findings first and include remediation guidance.
Inputs
- input (string, optional): Path to IAM policy JSON or raw JSON string for local mode.
- mode (string, optional): Execution mode: local, aws, or azure.
- dry_run (boolean, optional): Use bundled sample data instead of live API calls.
Bundled Files
- main.py: executable logic for the skill.
- README.md: detailed usage, input expectations, and sample outputs.
- sample_input: bundled example IAM policy inputs.
- skill.yaml: existing repo manifest retained for this project's original packaging.
Guidance
- Prefer local or dry-run analysis before live cloud access when possible.
- Prioritize wildcard actions, broad resource scope, and missing MFA or condition constraints.
- Clearly separate observed permissions from inferred risk.
