The Agent Skills Directory

[COMMAND_EXECUTION]: The skill provides instructions to execute shell commands using the crabbox CLI, such as crabbox job run detected and crabbox run --id <slug> -- <command>.
[INDIRECT_PROMPT_INJECTION]: The skill ingests and acts upon repository configuration files (crabbox.yaml or .crabbox.yaml), which are untrusted data sources. 1. Ingestion points: Detection of crabbox.yaml or .crabbox.yaml in the repository root. 2. Boundary markers: Absent; the skill relies on the agent's manual inspection of the configuration content. 3. Capability inventory: The crabbox CLI provides capabilities for remote shell command execution and SSH access. 4. Sanitization: No programmatic sanitization or schema validation of the configuration file is defined.

crabbox