The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill promotes the installation of a third-party utility named 'rtk' from external sources (rtk-ai/tap via Homebrew and rtk via Cargo). These sources and the 'rtk-ai' organization are not recognized as trusted vendors or well-known services, representing a supply chain risk.
[COMMAND_EXECUTION]: The skill instructs the agent to wrap standard shell commands (e.g., git log, cargo test, pytest, find) with the rtk binary. This allows an unverified external tool to process, filter, and potentially capture the output of sensitive development operations and local file data.
[INDIRECT_PROMPT_INJECTION]: The skill is designed to process and summarize high-verbosity data from external command outputs, which creates a surface for indirect prompt injection.
Ingestion points: Command outputs from git, find, pnpm, cargo, and pytest (SKILL.md).
Boundary markers: The skill lacks explicit instructions or delimiters to prevent the agent from being influenced by instructions embedded within the data processed by the rtk wrapper.
Capability inventory: The skill utilizes the Bash tool to execute commands and view file content.
Sanitization: There is no mechanism described to sanitize or validate the content returned by the commands before it is processed by the AI agent.

rtk-optimizer