The Agent Skills Directory

[DYNAMIC_EXECUTION]: The execution procedure involves generating and running temporary launch scripts to handle complex commands with special characters, which is a form of runtime script generation and execution.
[COMMAND_EXECUTION]: The skill relies on shell commands including tmux, lsof, kill, and curl to manage service lifecycles and detect live status. It executes arbitrary commands defined within the project's ctx/services.md manifest.
[PRIVILEGE_ESCALATION]: The skill provides instructions to install the tmux utility using system package managers like apt-get or brew if it is missing, which typically requires administrative or root privileges.
[INDIRECT_PROMPT_INJECTION]: The skill reads and acts upon instructions (start commands and session names) stored in the ctx/services.md file. If this file is modified by an external actor (e.g., via a malicious pull request), the agent may execute unintended commands.
Ingestion points: ctx/services.md manifest file.
Boundary markers: The skill uses a structured Markdown table for the manifest, but does not specify explicit boundary markers or 'ignore' instructions for the content of the commands.
Capability inventory: The skill has the capability to execute shell commands, manage background processes via tmux/setsid, and terminate processes via kill/lsof.
Sanitization: The skill includes a specific security constraint to avoid shell interpolation by building argument arrays or using temporary launch scripts.

ctx-serve