cloud-iam-deep

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 1.00). The skill includes examples and commands that embed API keys, service-account private keys, bearer tokens, and plaintext credentials (e.g., $API_KEY, sa_key['private_key'], Authorization: Bearer {token}, "minioadmin"), which requires handling and outputting secrets verbatim and thus poses high exfiltration risk.

CRITICAL E005: Suspicious download URL detected in skill instructions.

• Suspicious download URL detected (high risk: 0.90). Many of these URLs are object-storage and artifact-registry endpoints (S3/MinIO/Azure Blob/Google Storage/Artifact Registry/docker.pkg.dev) which are commonly used to host and distribute arbitrary files and container images — a high-risk vector for distributing malware if buckets/registries are public or writable.

CRITICAL E006: Malicious code pattern detected in skill scripts.

• Malicious code pattern detected (high risk: 1.00). This document is an explicit offensive playbook describing techniques to discover cloud project endpoints, abuse misconfigured serverless APIs, generate access tokens from stolen service-account keys, enumerate and download source/bucket/artifact data, create attacker accounts, and upload or delete data—demonstrating clear malicious intent to exfiltrate data and escalate access.

MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).

• Third-party content exposure detected (medium risk: 0.30). The skill’s required runtime workflow performs HTTP enumeration and method testing against Cloud Functions endpoints (e.g., /api/feed) by issuing requests to attacker-influenced URLs derived from project/region patterns and then ingesting returned free text (r.text, response bodies).