The Agent Skills Directory

[COMMAND_EXECUTION]: The skill instructs the agent to perform maintenance by running shell commands such as npm run sync, npm run inventory, and npm run validate. These commands execute scripts defined in the local repository environment. The skill also provides a list of agent-specific folders (e.g., .claude, .windsurf, .devin) to be removed during cleanup, which involves file system manipulation.
[EXTERNAL_DOWNLOADS]: The workflow relies on the npx skills add [owner/repo] command to fetch skill packs from arbitrary external GitHub repositories, which introduces third-party content into the workspace.
[REMOTE_CODE_EXECUTION]: Using npx to install external skill packages from user-provided GitHub repositories involves downloading and potentially executing code or configuration that defines those skills in the agent's execution environment.
[PROMPT_INJECTION]: By facilitating the ingestion of external skill files, the skill creates a surface for indirect prompt injection where instructions embedded in the imported content could influence the agent's behavior. 1. Ingestion points: External repositories retrieved via npx skills add as described in SKILL.md. 2. Boundary markers: Absent; there are no instructions provided to the agent to treat external content as untrusted or to use delimiters. 3. Capability inventory: The environment supports shell command execution (npm, npx) and recursive file deletion for cleanup tasks. 4. Sanitization: Absent; the skill does not require the agent to validate or sanitize external files before they are integrated into the canonical repository structure.

clean-skill-install