The Agent Skills Directory

[COMMAND_EXECUTION]: The skill instructions require the agent to execute specific CLI binaries like orca, orca-ide, and orca-dev based on environment detection.
[PRIVILEGE_ESCALATION]: The skill explicitly advises the agent to "re-run it with escalated permissions" when encountering sandbox-related access denials (runtime_access_denied), which encourages the agent to bypass security boundaries.
[DYNAMIC_EXECUTION]: The skill dynamically loads session-specific guides and instructions by executing ORCA skills get orca-cli at runtime, incorporating external content into its operation.
[INDIRECT_PROMPT_INJECTION]: The instruction loading mechanism creates an attack surface where external command output is directly integrated into the agent's context.
Ingestion points: Command output from ORCA skills get orca-cli in SKILL.md.
Boundary markers: Absent; the skill treats the output as a direct instruction guide.
Capability inventory: Full shell command execution capabilities via the Orca CLI suite.
Sanitization: No validation or filtering of the fetched content is performed before integration.

orca-cli