The Agent Skills Directory

[CREDENTIALS_UNSAFE]: The skill is configured to read from a local sensitive file path at ~/.config/zendesk/credentials.json to retrieve API tokens and user emails for authentication purposes.
[COMMAND_EXECUTION]: The skill relies on the execution of shell commands using curl and jq to interact with the Zendesk REST API and manipulate JSON data.
[PROMPT_INJECTION]: The skill has an indirect prompt injection surface because it fetches and processes ticket comments and search results which can contain arbitrary instructions from untrusted external users.
Ingestion points: Zendesk ticket comments, organization names, and user data fetched via the API (SKILL.md).
Boundary markers: None; data is processed directly as text strings or JSON objects without delimiters.
Capability inventory: Shell command execution via curl and jq across all documented operations (SKILL.md).
Sanitization: No sanitization or escaping of external content is performed before the data is passed to the agent context.

Zendesk