The Agent Skills Directory

[SAFE]: The skill provides instructional content for setting up the Koin DI framework. All external resources, including documentation links, Maven Central repository references, and Kotzilla service endpoints, are legitimate and belong to the vendor's (insertkoinio) ecosystem or well-known technology services.
[INDIRECT_PROMPT_INJECTION]: As outlined in Step 0 of SKILL.md, the agent reads local project build files and source code to determine the project type and existing DI configuration. While this process ingests untrusted data from the user's local environment, the skill limits the use of this information to branching logic for setup steps and does not execute instructions embedded within those files. The risk is assessed as safe given the intended purpose and lack of dangerous interpolation of this data into high-privilege commands.
[REMOTE_CODE_EXECUTION]: Instructions in references/kotzilla.md guide the user to manually add an MCP server for runtime monitoring. This is a transparent instruction for the user to configure their environment with a known tool associated with the Koin ecosystem and contains explicit rules against silent or automatic installation by the agent.

koin-setup