The Agent Skills Directory

[REMOTE_CODE_EXECUTION]: The skill instructs the agent to execute code directly from a remote repository using npx -y github:timschoch/skilly. While this is a vendor-owned resource, executing code directly from a remote source at runtime is a high-risk pattern.
[REMOTE_CODE_EXECUTION]: The instructions allow the agent to install skills from arbitrary URLs provided by the user via npx -y skills add <url>. This provides a direct mechanism for the agent to download and execute potentially malicious or unverified code from the internet.
[COMMAND_EXECUTION]: The skill directs the agent to perform multiple shell commands, including npx for package execution and git operations such as git branch -m for branch management.
[INDIRECT_PROMPT_INJECTION]: The skill has a significant attack surface for indirect prompt injection due to its core functionality of processing external data.
Ingestion points: User-provided URLs intended for skill installation via the skills add command in SKILL.md.
Boundary markers: No explicit boundary markers or instructions to ignore embedded directives are provided for the content fetched from external URLs.
Capability inventory: The skill possesses the capability to execute shell commands (npx), modify the local filesystem, and interact with git repositories (commits, pushes, and branch management) as described in SKILL.md.
Sanitization: There is no evidence of URL validation or content sanitization before the external resource is processed and added to the agent's environment.

skilly-cli