The Agent Skills Directory

[COMMAND_EXECUTION]: The skill instructions and verification scripts rely on the execution of the cloudrobo CLI to interact with cloud services.
Evidence: Core command definitions in SKILL.md and functional tests in scripts/test-cli-commands.sh invoke the cloudrobo resource command group.
[EXTERNAL_DOWNLOADS]: The skill documentation guides the user to install the hw-cloudrobo-client package, which is an official resource provided by the vendor.
Evidence: references/cli-installation-guide.md provides installation instructions for the hw-cloudrobo-client package via pip.
[INDIRECT_PROMPT_INJECTION]: The skill processes potentially untrusted data retrieved from cloud APIs (such as resource names, node labels, and descriptions) which could be used as an injection vector if the cloud environment is compromised.
Ingestion points: Output from list-quotas, list-pools, and show-pool commands as described in SKILL.md.
Boundary markers: Absent; the agent is expected to process the raw output from CLI/SDK calls.
Capability inventory: The skill utilizes subprocess execution for the CLI and Python environment for the SDK.
Sanitization: Absent; the skill does not explicitly describe sanitization of the cloud-provided metadata before presentation to the agent.
[DYNAMIC_EXECUTION]: The provided functional test script generates and runs short Python scripts to verify SDK behavior at runtime.
Evidence: scripts/test-cli-commands.sh uses python3 -c to dynamically execute code snippets that instantiate the ResourceClient and call its methods.

huawei-cloud-cloudrobo-resource