The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill fetches the bg management utility from the author's official GitHub repository at github.com/lirrensi/agent-sommelier during installation.
[REMOTE_CODE_EXECUTION]: The installation process involves downloading and executing a remote tool from GitHub via the uv tool manager.
[COMMAND_EXECUTION]: The primary purpose of the skill is to execute arbitrary shell commands in the background using bg run. While this is intended behavior, it provides a powerful interface for system interaction.
[PROMPT_INJECTION]: The skill is vulnerable to indirect prompt injection by ingesting and returning output from external processes to the agent.
Ingestion points: Captured stdout and stderr from background processes located in the agentcli_bgjobs/records/ directory.
Boundary markers: No delimiters or safety instructions are used when reading job output via bg read or bg logs.
Capability inventory: The skill possesses full shell command execution capabilities through bg run and file reading capabilities through bg read.
Sanitization: The skill does not perform any sanitization, filtering, or escaping of the output generated by background tasks before it is presented to the agent context.

bg-jobs