The Agent Skills Directory

[SAFE]: The skill instructions and associated Python script do not contain any malicious code, obfuscation, or persistence mechanisms. All logic is focused on legitimate Kotlin Multiplatform project configuration and migration.
[INDIRECT_PROMPT_INJECTION]: The skill analyzes external code and configuration files, which presents a standard surface for indirect prompt injection common to all code analysis tools.
Ingestion points: The scripts/inspect-project.py script reads project files including .kt, .gradle.kts, .properties, and .toml files from the repository root as part of the structure inspection in Step 2 of SKILL.md.
Boundary markers: The inspection script returns JSON metadata to the agent without specific delimiters or instructions to ignore embedded content in the scanned files.
Capability inventory: The skill allows the agent to read repository files and execute the local inspect-project.py script; it does not grant network access or file-writing permissions outside the repository context.
Sanitization: The script extracts project-specific strings (like dependency names and target configurations) without escaping, though these are used for project diagnostics rather than direct shell execution.

kotlin-multiplatform