The Agent Skills Directory

[COMMAND_EXECUTION]: The skill relies on the lx CLI to perform administrative tasks on knowledge entries and files. It also utilizes curl as a necessary step in the file upload workflow to transmit data to a pre-signed URL. \n- [EXTERNAL_DOWNLOADS]: The skill provides tools to retrieve download links for files stored in the knowledge base via the lx file download-file command.\n- [PROMPT_INJECTION]: The skill has an attack surface for indirect prompt injection because it reads and parses external content from the knowledge base (e.g., through lx entry describe-ai-parse-content) to provide information to the agent.\n
Ingestion points: The lx entry describe-ai-parse-content command retrieves HTML, Markdown, or OCR content from specified entries (documented in SKILL.md and entry-crud.md).\n
Boundary markers: The instructions do not specify any delimiters or warnings to ignore instructions found within the retrieved entry content.\n
Capability inventory: The skill possesses significant write and modification capabilities, such as create-entry, move-entry, import-content, set-entry-tags, and save-markdown-draft.\n
Sanitization: There is no evidence of content sanitization or validation before the retrieved data is processed by the agent.

lx-entry