The Agent Skills Directory

[COMMAND_EXECUTION]: The skill instructs the agent to execute vendor-specific CLI commands to authenticate the user and process tasks. • Evidence: Instructions to run npx @kodycodes/cli auth bootstrap and npx @kodycodes/cli whoami (SKILL.md).
[INDIRECT_PROMPT_INJECTION]: The skill establishes a surface for processing external data retrieved via tool outputs. • Ingestion points: The search tool is used to ingest data from external packages, integrations, and documentation (SKILL.md). • Boundary markers: No specific delimiters or safety warnings are provided for search results. • Capability inventory: The skill utilizes execute and CLI tools to run code based on retrieved data (SKILL.md). • Sanitization: No explicit sanitization of search results is mentioned.
[DYNAMIC_EXECUTION]: The workflow relies on the execution of code derived from external resources or packages at runtime. • Evidence: Instructions to run packages through execute or the local CLI and to persist work as reusable Kody packages (SKILL.md).

use-kody