The Agent Skills Directory

[DATA_EXFILTRATION]: The skill accesses the local file ~/.uipath/.auth and environment variables such as UIPATH_CLI_AUTH_TOKEN to retrieve session tokens. These credentials are used to authenticate requests to official UiPath cloud endpoints (e.g., cloud.uipath.com). This behavior is consistent with the primary purpose of managing UiPath resources and follows standard authentication patterns for the vendor's tooling.
[INDIRECT_PROMPT_INJECTION]: The skill reads external data from Automation Hub, including process metadata and user-contributed comments. This data is then summarized or processed, creating a surface for potential indirect prompt injection attacks.
Ingestion points: Data is ingested via the Automation Hub Open API endpoints (/automations/{id}, /automations/{id}/comments) or the uip CLI tool.
Boundary markers: The instructions do not define specific delimiters or warnings for the agent when processing retrieved process data or comments.
Capability inventory: The skill utilizes shell tools (Bash) for command execution (uip, curl) and file system operations within the workspace.
Sanitization: No explicit sanitization or filtering of the retrieved content is specified before it is presented to the agent context.

uipath-automationhub