The Agent Skills Directory

[COMMAND_EXECUTION]: The skill provides the cmux send and cmux send-key commands, which allow the agent to execute arbitrary shell commands within managed terminal surfaces. This is a core feature for terminal orchestration.
[COMMAND_EXECUTION]: The skill enables arbitrary JavaScript execution within browser surfaces using the cmux browser eval command, which is used for advanced browser automation tasks.
[DATA_EXFILTRATION]: The skill includes capabilities to read terminal output (cmux read-screen) and browser secrets (cmux browser cookies get). Additionally, the cmux pipe-pane command can write terminal output history to a specified file, which could involve sensitive data if not managed carefully by the agent.
[PROMPT_INJECTION]: The skill ingests untrusted data from terminal outputs and web pages, creating an attack surface for indirect prompt injection.
Ingestion points: Terminal output is ingested via cmux read-screen (documented in SKILL.md), and web content is ingested via cmux browser snapshot and cmux browser get html (documented in references/browser-automation.md).
Boundary markers: Absent. The instructions do not provide delimiters or warnings to the agent regarding instructions that might be embedded in the read terminal or browser content.
Capability inventory: The skill possesses powerful execution capabilities, including shell command execution (cmux send) and browser script execution (cmux browser eval).
Sanitization: Absent. There are no instructions for validating or escaping external content before it is processed by the agent.

cmux