The Agent Skills Directory

[PRIVILEGE_ESCALATION]: The skill explicitly instructs the agent to re-run commands with escalated permissions (e.g., sudo or administrative rights) if it encounters a runtime_access_denied error. This encourages the agent to bypass operating system security boundaries.
[DYNAMIC_EXECUTION]: The skill uses the ORCA skills get command to fetch and load additional instructions (guides) at runtime. This introduces dynamic content into the agent's context that is not visible during static analysis of the skill file.
[COMMAND_EXECUTION]: The skill provides logic for the agent to resolve and execute local binaries (orca, orca-ide, orca-dev) based on environment variables like ORCA_CLI_COMMAND and ORCA_DEV_REPO_ROOT, or specific system configurations.
[INDIRECT_PROMPT_INJECTION]: The skill has an attack surface for indirect injection by ingesting external data through the ORCA skills get command.
Ingestion points: The ORCA skills get orca-emulator-android command in SKILL.md fetches external content into the session.
Boundary markers: None are specified to delimit the guide content from the main instructions.
Capability inventory: The skill allows Android device control, app installation, and interaction with the accessibility tree via adb and the Orca CLI.
Sanitization: There is no mention of sanitizing or validating the content returned by the skills get command.

orca-emulator-android