The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill documentation includes instructions to install the omniroute package from the NPM registry. This package is the core utility for the service the skill supports.
[COMMAND_EXECUTION]: The skill's functionality is entirely based on executing omniroute CLI subcommands. These commands allow the agent to interact with the OmniRoute service to manage credentials and authentication flows.
[INDIRECT_PROMPT_INJECTION]: The skill has an attack surface for indirect prompt injection because it reads and processes the output of CLI commands, which could theoretically contain malicious instructions if the upstream service or local environment were compromised.
Ingestion points: The agent parses output from subcommands such as keys list and oauth status into its context.
Boundary markers: No specific delimiters or safety instructions are provided to the agent for handling the CLI output.
Capability inventory: The skill allows for various CLI operations including key creation, rotation, and revocation.
Sanitization: The instructions do not define any validation or sanitization logic for the data returned by the CLI tool.

cli-keys