aliyun-ecs
Pass
Audited by Gen Agent Trust Hub on Sep 14, 2026
Risk Level: SAFECREDENTIALS_UNSAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [CREDENTIALS_UNSAFE]: The skill documents procedures for retrieving temporary security credentials (STS tokens) from the Alibaba Cloud Metadata Service at http://100.100.100.200/latest/meta-data/ram/security-credentials/. This allows access to sensitive information such as AccessKeyId, SecretAccessKey, and SecurityToken for assigned RAM roles.\n- [COMMAND_EXECUTION]: The documentation in references/authentication.md describes the External authentication mode of the aliyun CLI, which is designed to execute external programs or commands to dynamically generate credentials. This feature provides a mechanism for arbitrary shell execution if the agent is manipulated into using untrusted commands.\n- [COMMAND_EXECUTION]: The skill includes setup instructions using yum install aliyun-cli -y, which requires administrative privileges to install system-level software.\n- [INDIRECT_PROMPT_INJECTION]: The skill processes data from external cloud APIs, such as instance tags and metadata, which are attacker-controllable. This creates a vulnerability surface where an attacker could inject malicious instructions into resource attributes.\n
- Ingestion points: API commands like aliyun ecs DescribeInstances in SKILL.md that retrieve resource metadata.\n
- Boundary markers: Absent; there are no instructions for the agent to use delimiters or ignore instructions embedded in the retrieved resource data.\n
- Capability inventory: The skill provides a wide range of powerful management commands, including RunInstances, StopInstance, and DeleteInstance in SKILL.md, which could be abused if an injection is successful.\n
- Sanitization: Absent; the instructions do not include steps for validating or filtering the content returned from the cloud APIs.
Audit Metadata