The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses the Bash tool to execute parallel-cli commands. These operations are restricted to the parallel-cli:* scope, which follows the principle of least privilege for the tool execution environment.
[INDIRECT_PROMPT_INJECTION]: The skill processes data from external websites during monitoring activities. This represents a potential attack surface if monitored content contains instructions designed to influence the agent's behavior during summarization or event reporting.
Ingestion points: External web content is ingested through the parallel-cli monitor events command output (SKILL.md).
Boundary markers: None explicitly defined in the prompt instructions.
Capability inventory: File system access and network operations are mediated through the parallel-cli tool.
Sanitization: The skill relies on the agent's default processing and the CLI's internal handling of web data.
[DATA_EXFILTRATION]: The skill supports a --webhook flag, which allows the agent to send monitoring event data to an external URL. While this is a documented feature for automated alerts, it creates a mechanism for sending web-derived data to third-party endpoints.

parallel-monitor