The Agent Skills Directory

[COMMAND_EXECUTION]: The skill makes extensive use of system commands including gh (GitHub CLI) and git to automate repository forking, cloning, and configuration. These operations are essential for the skill's stated purpose of spawning new instances.
[DATA_EXPOSURE]: The skill reads and modifies local metadata and state files, specifically memory/instances.json and memory/MEMORY.md, to maintain a registry of deployed instances and their purposes.
[PROMPT_INJECTION]: The skill processes a user-supplied variable ${var} which contains the name and purpose of the new instance. While the repository name is sanitized, the 'purpose' string is interpolated directly into the new instance's core documentation (CLAUDE.md, SETUP.md, and MEMORY.md). This creates an indirect prompt injection surface where a malicious purpose string could influence the behavior of the newly spawned agent instance.
Ingestion points: User-provided ${var} variable in SKILL.md.
Boundary markers: No explicit delimiters are used when interpolating the purpose into markdown documentation files.
Capability inventory: The skill has capabilities for shell command execution (gh, git), filesystem writes, and repository configuration.
Sanitization: The NAME component is strictly sanitized to lowercase alphanumeric characters; however, the PURPOSE component is only trimmed and not validated against instructional content.

spawn-instance