The Agent Skills Directory

EXTERNAL_DOWNLOADS (HIGH): The skill's core functionality is to fetch content from external sources defined in assets/skill-sources.yaml. The primary source listed (https://github.com/kehwar/frappe) is not a trusted organization or repository, violating the trust-scope-rule.
REMOTE_CODE_EXECUTION (HIGH): The imported 'skills' are intended to be executed or interpreted by the AI agent. By facilitating the download of these scripts and markdown instructions into the local .github/skills/ directory without verification, the skill enables a Remote Code Execution (RCE) vector.
COMMAND_EXECUTION (MEDIUM): The documentation in SKILL.md provides explicit instructions for the agent to execute shell commands, including git clone, cp -r, and running a local Python script (scripts/import_skills.py) to modify the local file system.
INDIRECT PROMPT INJECTION (LOW): The skill creates an attack surface for indirect prompt injection by processing data from untrusted external repositories.
Ingestion points: Remote GitHub repositories listed in assets/skill-sources.yaml.
Boundary markers: Absent; there are no delimiters or warnings to ignore instructions within the imported content.
Capability inventory: The skill possesses the ability to write to the local file system and execute shell commands.
Sanitization: Absent; the process involves a direct copy of remote files into the local skill directory.

skill-importer