The Agent Skills Directory

[SAFE]: The skill operates entirely through the official @uipath/cli tool, ensuring all operations are performed via validated vendor-supported channels.
[COMMAND_EXECUTION]: The skill utilizes the uip solution CLI for essential lifecycle operations such as init, pack, publish, and deploy. These commands are standard for infrastructure-as-code and automation management workflows.
[DYNAMIC_CONTEXT_INJECTION]: The skill employs the !uip solution init --help --output json pattern in its instructions to detect the CLI version. This is a benign and functional use of dynamic execution to ensure compatibility with different versions of the UiPath CLI.
[EXTERNAL_DOWNLOADS]: The documentation correctly points users to the official npm registry to install the @uipath/cli package. This is a standard and safe delivery mechanism for developer tooling.
[INDIRECT_PROMPT_INJECTION]: The skill identifies a potential attack surface where external project metadata files (.uipx, bindings_v2.json, .uis) are ingested to inform CLI commands.
Ingestion points: Metadata files (.uipx, project.json, bindings_v2.json) and exported solution archives (.uis).
Boundary markers: Explicit boundary markers are not present in the files themselves, but the skill relies on the CLI's internal parsing logic.
Capability inventory: Subprocess execution via uip solution commands and resource spec modification via uip solution resources edit --patch.
Sanitization: The skill delegates input sanitization and schema validation to the underlying uip CLI tool, which is designed to handle these data formats safely.

uipath-solution