The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill configuration specifies the installation of a binary from a third-party Homebrew repository (steipete/tap/summarize).
[COMMAND_EXECUTION]: The skill provides instructions for executing the summarize CLI tool via shell commands to process URLs and file paths.
[PROMPT_INJECTION]: The skill has a surface for indirect prompt injection because it is designed to ingest and process untrusted external data from URLs and YouTube transcripts.
Ingestion points: Content is pulled from external URLs, YouTube transcripts, and local files provided by the user (SKILL.md).
Boundary markers: The current command structure does not specify delimiters to separate external content from system instructions.
Capability inventory: The skill executes shell commands and makes network requests to external LLM providers (OpenAI, Anthropic, Google, xAI).
Sanitization: No sanitization or filtering of the external input content is specified before it is processed by the model.

summarize