The Agent Skills Directory

[COMMAND_EXECUTION]: The skill facilitates the execution of the b2c CLI tool to perform site administrative tasks. This includes commands for importing site archives (b2c job import), managing import sets (b2c job import-set), and interacting with remote file storage via WebDAV (b2c webdav rm). These operations are aligned with the skill's stated purpose of managing Salesforce B2C Commerce instances.\n- [CREDENTIALS_UNSAFE]: The skill interacts with sensitive configuration files and environment variables used for authentication, such as SFCC_* variables, dw.json, and ~/.mobify. It also provides instructions for using b2c setup inspect --unmask to reveal stored secrets. While these are necessary for the tool's function, they involve the handling of high-value credentials. The skill includes explicit warnings to remove sensitive data from exported archives before committing them to version control.\n- [INDIRECT_PROMPT_INJECTION]: The skill is designed to process external data in the form of site archive directories and metadata XML files (e.g., system-objecttype-extensions.xml). This ingestion of untrusted local data represents a surface for indirect prompt injection.\n
Ingestion points: Local directory paths and XML files provided to b2c job import and b2c job import-set (found in SKILL.md).\n
Boundary markers: The instructions do not define specific delimiters or warnings to prevent the agent from interpreting instructions embedded within the metadata XML.\n
Capability inventory: The skill utilizes the b2c CLI, which has capabilities for file system access, network communication with B2C Commerce APIs, and job execution (found in SKILL.md).\n
Sanitization: There is no mention of automated validation or sanitization of the contents of the imported XML files beyond standard schema validation.

b2c-site-import-export