The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses the uip CLI tool via the Bash tool to perform connector management, validation, and publishing. All commands are consistent with the stated purpose of authoring connectors for the UiPath platform.
[DYNAMIC_EXECUTION]: The skill provides patterns and instructions for writing JavaScript hooks to be executed by the UiPath Integration Service "Denali" engine at runtime. The instructions include security-focused examples, such as validating dynamic URLs using host allowlists and protocol checks.
[INDIRECT_PROMPT_INJECTION]: The skill defines a surface for indirect prompt injection as it instructs the agent to process external vendor API documentation and generate connector configurations and hooks.
Ingestion points: Vendor API documentation read from local files or provided in prompts.
Boundary markers: The skill relies on the builder validate tool to identify schema errors or missing fields.
Capability inventory: The agent can execute uip shell commands, write connector files (JSON and JS), and publish the resulting connectors to a live tenant.
Sanitization: The skill explicitly provides patterns for input validation in hooks, such as host suffix allowlisting for dynamically derived base URLs.
[SAFE]: The skill is authored by 'uipath' and relates to official 'uipath' tooling. It explicitly warns against hardcoding or logging secrets (Rule 9) and follows standard secret management practices for the platform.

uipath-connector-builder