The Agent Skills Directory

[COMMAND_EXECUTION]: The skill is designed specifically to execute shell commands related to the Hydra node and Cardano CLI. It utilizes command-dispatch: tool with disable-model-invocation: true, which is a security feature to ensure that the AI model cannot autonomously generate or execute commands; instead, it acts as a gateway for user-provided commands.
[EXTERNAL_DOWNLOADS]: The scripts/hydra-node.sh script executes docker run using an image from ghcr.io/cardano-scaling/hydra-node. This is an official repository for the Cardano Hydra project and is considered a safe, well-known source for this domain.
[DYNAMIC_CONTEXT_INJECTION]: The SKILL.md frontmatter contains dynamic context markers (!hydra-node --version and !hydra-node run --help) used to populate the agent's context with version and help information at load time. These are benign utility commands used for environment discovery.
[DATA_EXPOSURE]: The skill provides instructions for generating and handling sensitive cryptographic keys (hydra.sk, cardano.sk). It correctly advises the user to set restrictive permissions (chmod 600) and warns never to share these files.

hydra-head-operator