The Agent Skills Directory

[REMOTE_CODE_EXECUTION]: The skill provides instructions for installing the UiPath CLI by downloading a script from download.uipath.com and piping it directly to the shell (curl -fsSL https://download.uipath.com/uipath-cli/install.sh | bash). This matches the standard installation pattern for the vendor's official tooling.
[EXTERNAL_DOWNLOADS]: Fetches installation scripts (install.sh, install.ps1) from the uipath.com domain.
[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data, including user conversation history and local project files, to construct feedback reports.
Ingestion points: Conversation history (prompts) and project files (e.g., .flow, project.json, pyproject.toml) are read into the agent's context.
Boundary markers: The skill does not explicitly define delimiters for the captured content within the feedback description, though it uses structured headers.
Capability inventory: The skill uses Bash for command execution, Write for file management, and uses the uip CLI to send data over the network.
Sanitization: The skill implements detailed sanitization rules in the 'Sanitization Rules' section, including stripping secrets (tokens, passwords), redacting PII in paths, and truncating large files to mitigate data exposure risks.

uipath-feedback