The Agent Skills Directory

[COMMAND_EXECUTION]: The skill utilizes the docyrus CLI to perform administrative tasks on brand records. Commands such as docyrus account brands create, update, and delete are used to manage tenant-scoped data. This execution is limited to the vendor's own platform tooling and is necessary for the skill's functionality.
[EXTERNAL_DOWNLOADS]: The fetch-from-website command enables the agent to scrape content from a brand's website to automatically populate identity fields. This involves retrieving data from user-provided external URLs.
[PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface through its data ingestion features.
Ingestion points: Data is ingested from external websites via fetch-from-website and from existing brand records via get and list commands as documented in SKILL.md.
Boundary markers: No specific delimiters or instructions to ignore embedded commands are applied to the data retrieved from the CLI or scraped from websites.
Capability inventory: The agent can execute docyrus CLI commands to modify tenant brand records, which could be influenced if the ingested data contains malicious instructions.
Sanitization: The skill documentation does not specify validation or sanitization of ingested content (such as brand voice guidelines or scraped text) before it is processed.

docyrus-tenant-brand-management