The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill ingests requirement descriptions and test case data from external Test Manager projects to drive AI-based generation and evaluation tasks.
Ingestion points: Requirement descriptions are fetched via uip tm requirements get and list commands as described in references/requirement-ai-guide.md.
Boundary markers: The instructions do not define explicit boundary markers or delimiters when requirement text is interpolated into AI prompts.
Capability inventory: The agent has the capability to execute shell commands (uip, playwright), write files (reports), and modify Test Manager resources.
Sanitization: There is no explicit sanitization of fetched text; however, the skill mitigates risk by requiring human-in-the-loop validation, where the agent must present generated test cases to the user for approval before creation.
[COMMAND_EXECUTION]: The skill is primarily a wrapper for the uip CLI tool. It utilizes the Bash tool to execute a wide variety of management commands. Security is maintained through 'Critical Rule 10', which instructs the agent to stop and consult the user on any failure or ambiguity during command execution.
[DYNAMIC_EXECUTION]: During Playwright suite processing, the skill executes playwright test --list for test discovery. This involves running code within the local project directory to identify test metadata. This is a standard and expected behavior for the described workflow.

uipath-test