The Agent Skills Directory

[COMMAND_EXECUTION]: The skill utilizes several shell commands including curl, git, grep, and cat to perform deployment tasks. These commands are used to query API versions, clone repositories, and modify configuration files as part of the intended GitOps workflow.
[EXTERNAL_DOWNLOADS]: The skill fetches data and clones repositories from code.choerodon.com.cn and chart.choerodon.com.cn. These are recognized as official infrastructure for the YQCloud service being managed.
[PROMPT_INJECTION]: The skill presents an indirect prompt injection surface as it ingests and processes JSON data from the ChartMuseum API and YAML configuration files from Git repositories.
Ingestion points: Data enters the context via curl output saved to ~/tmp/ and git clone operations on deployment repositories.
Boundary markers: None identified in the provided instructions.
Capability inventory: Includes filesystem access (mkdir, cd, tee), network access (curl), and Git operations (git push, git commit).
Sanitization: No explicit sanitization or validation of the fetched JSON or YAML content is performed before processing.
[CREDENTIALS_UNSAFE]: The documentation suggests the possibility of passing credentials via the -u username:password flag in curl commands. While no hardcoded credentials are provided, this practice can lead to the exposure of secrets in environment logs or command history.

yqcloud-deploy