The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill is designed to fetch external content, including scripts (Python, shell) and configuration files, from the prompts.chat MCP server.
[COMMAND_EXECUTION]: The instructions direct the agent to perform file system operations, such as creating directories and writing files to .claude/skills/ on the local machine.
[PROMPT_INJECTION]: The skill processes untrusted external data, creating a surface for indirect prompt injection.
Ingestion points: Data returned by the search_skills and get_skill MCP tools (SKILL.md, metadata, and script contents).
Boundary markers: Absent; there are no instructions to delimit fetched data or ignore embedded instructions within the downloaded skills.
Capability inventory: File system write access and directory creation are used to store the retrieved content.
Sanitization: Absent; the skill does not specify any validation, escaping, or filtering of the content retrieved from the marketplace.

skill-lookup