The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data from multiple sources within the Aliyun Yunxiao ecosystem which could contain malicious prompts.
Ingestion points: Reads repository file blobs (references/code-management.md), work item descriptions, and user comments (references/project-management.md).
Boundary markers: None identified in the provided instructions; external content appears to be processed without explicit delimiters or safety warnings.
Capability inventory: Significant capabilities including writing to code repositories, creating branches, executing CI/CD pipelines, and modifying project management items.
Sanitization: No evidence of sanitization or filtering of harvested external content before it is processed by the agent.
[COMMAND_EXECUTION]: The skill involves local system commands for metadata gathering.
Evidence: Documentation in references/pipeline.md specifies execution of local Git commands such as git config --get remote.origin.url and git branch --show-current to automate pipeline creation.

yunxiao-devops