The Agent Skills Directory

[COMMAND_EXECUTION]: The skill provides the AI agent with a specific decision tree to execute shell commands via the cs CLI tool. This includes modifying local configurations (cs config set), performing diagnostics (cs doctor), and resetting the environment (cs --setup), which allows the agent to alter the local system state based on user requests.
[REMOTE_CODE_EXECUTION]: The skill explicitly instructs the agent to use the cs upgrade command to update the tool. This process triggers a remote update mechanism controlled by the cs binary, which downloads and executes new code versions from external sources to maintain the software.
[INDIRECT_PROMPT_INJECTION]: The skill contains a potential attack surface for indirect prompt injection.
Ingestion points: The skill references local status files, specifically ~/.agentparty/state/<workspaceId>/statusline.json, which contain message previews and activity data.
Boundary markers: No explicit delimiters or instructions to ignore embedded commands are defined for the content of these processed files.
Capability inventory: The agent has the capability to execute shell commands and modify configurations based on interpreted data.
Sanitization: There is no mention of validating or escaping the contents of the external JSON files before they are presented to the agent, which could allow maliciously crafted local data to influence the agent's behavior.
[DATA_EXPOSURE]: The skill is designed to read and modify sensitive local configuration files, such as ~/.claude/settings.json, to manage the status bar integration. While functional, this represents high-privilege access to the user's application environment settings.

claude-statusbar