- name
- cloud-overview
- description
- Cloud exploitation lane — AWS IAM privesc, S3 takeover, k8s RBAC abuse, Terraform state leaks, cloud metadata pivoting.
- metadata
- {"subdomain":"cloud","when_to_use":"cloud aws gcp azure iam s3 kubernetes k8s terraform metadata imds privesc lane overview routing","mitre_attack":["T1078.004","T1530","T1552.005","T1552.007","T1610","T1611"]}
# Cloud Hunter Skill Catalog
## Playbooks
| Skill | Use for |
|---|---|
| `/skills/standard/cloud/aws-iam-enum/SKILL.md` | IAM enumeration + privesc |
| `/skills/standard/cloud/s3-takeover/SKILL.md` | Dangling bucket / subdomain takeover |
| `/skills/standard/cloud/k8s-pivot/SKILL.md` | Pod escape, RBAC abuse, hostPath |
| `/skills/standard/cloud/terraform-state-leak/SKILL.md` | Exposed state file exploitation |
| `/skills/standard/cloud/imds-pivot/SKILL.md` | SSRF → metadata → IAM role |
## Workflow (authenticated engagement)
1. `bash("aws sts get-caller-identity")`
2. `bash("aws iam list-attached-user-policies --user-name <me>")`
3. For each attached policy: fetch JSON and `iam_policy_audit`
4. Feed Terraform state via `bash("aws s3 cp s3://bucket/terraform.tfstate -")` → `tfstate_audit`
5. `bash("kubectl get pods -A -o json")` → `k8s_audit`
6. Every privesc primitive → kg_add_node + chain edges
## Workflow (post-SSRF)
1. `metadata_endpoints("aws")` for the target cloud
2. Pivot URL one at a time via the SSRF vector
3. Confirmed creds → `credential` node + `leaks` edge from the SSRF vuln
4. `plan_attack_chains(promote=True)` to see the full path
GitHub에서 보기