The Agent Skills Directory

[COMMAND_EXECUTION]: The skill utilizes the Bash tool to interact with the uip admin CLI. This allows it to perform management tasks across user identities, organizational tenants, and security configurations.
[INDIRECT_PROMPT_INJECTION]: The skill processes data returned from the UiPath platform via CLI commands.
Ingestion points: External data is fetched using uip admin commands in various reference files (e.g., audit-workflow-guide.md, user-management.md).
Boundary markers: The skill uses structured JSON output and jq for parsing, providing logical separation between command data and instructions.
Capability inventory: The skill has the ability to manage users, roles, IP restrictions, and write audit exports to the file system.
Sanitization: Instructions are included to sanitize CSV exports against formula injection and to validate file paths to prevent traversal attacks.
[CREDENTIALS_UNSAFE]: The skill handles the creation and management of credentials like OAuth2 secrets and personal access tokens. It follows security best practices by instructing the agent to show secrets only once and advising the user to store them immediately.
[PRIVILEGE_ESCALATION]: The skill can modify administrative roles and permissions. It includes a security protocol requiring the resolution of principal identities and explicit user confirmation before any modifications are executed.

uipath-admin