The Agent Skills Directory

[COMMAND_EXECUTION]: The skill resolves and executes local binaries (specifically orca, orca-ide, or orca-dev) to manage system state. It prioritizes the ORCA_CLI_COMMAND environment variable and includes logic to prevent execution of the GNOME screen reader on Linux systems.
[INDIRECT_PROMPT_INJECTION]: The skill employs a dynamic instruction-loading pattern where the agent is directed to fetch its full operational guide from the output of ORCA skills get orca-cli. This creates a surface where the binary's output dictates subsequent agent behavior.
Ingestion points: The shell output from the orca executable as described in SKILL.md.
Boundary markers: Absent; the agent is instructed to read the output as a version-matched guide for its next commands.
Capability inventory: The skill manages terminals, worktrees, and browsers, providing significant interaction capabilities with the local environment.
Sanitization: Absent; the skill trusts the output of the resolved local binary as authoritative instructions.

orca-cli