alibabacloud-cas-ssl-common-tools

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 1.00). The skill template instructs the agent to echo/embed private key and certificate PEM contents and to place passwords into shell command arguments (e.g., echo "{{key_content}}", --Key "$(cat ...)", and convert-format.sh "{{password}}"), which requires the LLM to handle and output secret values verbatim.

CRITICAL E006: Malicious code pattern detected in skill scripts.

• Malicious code pattern detected (high risk: 0.90). The toolkit contains explicit instructions to create a long-lived IAM role that trusts an external service principal (aideepsign.aliyuncs.com) and to attach broad system policies (cas:* / alidns:*), which can grant persistent remote access (a backdoor / supply-chain escalation) to the account—this is a high-risk, potentially abusive pattern.

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

• Third-party content exposure detected (medium risk: 0.30). The runtime path for “Domain Verify / Certificate Download” ingests free-text fields from a cloud API response (ValidationValue/FileContent) via get-instance-detail and then incorporates them into commands/output, so an outsider that can create/affect those certificate verification/app data can poison what the workflow reads.

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

• Potentially malicious external URL detected (high risk: 1.00). The skill's installation guide includes runtime download-and-execute instructions (curl/Invoke-WebRequest) for the Aliyun CLI from https://aliyuncli.alicdn.com/setup.sh and GitHub release URLs (e.g., https://github.com/aliyun/aliyun-cli/releases/download/v${VERSION}/aliyun-cli-linux-${VERSION}-amd64.tgz and related SHASUMS files), which fetch remote binaries/scripts that may be executed during runtime.