hunt-cloud-misconfig

CRITICAL E004: Prompt injection detected in skill instructions.

• Potential prompt injection detected (critical risk: 1.00). This skill contains explicit, actionable instructions for discovering and exploiting cloud misconfigurations (S3/GCS/Azure buckets, Firebase write, EC2 metadata via SSRF) including methods to retrieve cloud credentials and perform read/write/exfiltration.

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

• Third-party content exposure detected (medium risk: 0.30). The skill “hunt-cloud-misconfig” ingests outsider-authored free text at runtime via monitoring/consuming user-supplied targets and performing discovery/enumeration (e.g., dorking/certificate-CT/port scans and then reading from public misconfig endpoints like public S3/GCS/Azure panels), which can cause the agent to fetch and process arbitrary page content or responses provided indirectly by the outsider.

MEDIUM W012: Unverifiable external dependency detected (runtime URL that controls agent).

• Potentially malicious external URL detected (high risk: 0.70). The skill's local verification instructions include a runtime command that pulls and runs a remote container image (localstack/localstack:3.0), which fetches and executes remote code locally.