The Agent Skills Directory

[PRIVILEGE_ESCALATION]: The skill instructions explicitly direct the agent to re-run commands with escalated permissions (typically involving sudo or Administrator rights) if a runtime_access_denied error occurs during execution.
[COMMAND_EXECUTION]: The skill requires the agent to execute local binaries (orca, orca-ide, or orca-dev) determined by the presence of environment variables like ORCA_CLI_COMMAND and ORCA_DEV_REPO_ROOT, creating a dependency on the local execution environment's integrity.
[DYNAMIC_EXECUTION]: The skill uses dynamic logic to resolve an executable path at runtime and then uses that resolved path to fetch and load additional instructions or content into the session.
[INDIRECT_PROMPT_INJECTION]: The skill fetches and processes a "version-matched guide" via the command ORCA skills get orca-per-workspace-env, creating an attack surface where external data from a local binary is ingested as agent instructions.
Ingestion points: Output of the ORCA skills get command (SKILL.md)
Boundary markers: None present to distinguish between static instructions and fetched content.
Capability inventory: Shell command execution via the ORCA placeholder and suggested privilege escalation.
Sanitization: No sanitization or validation of the fetched content is specified before execution.

orca-per-workspace-env