The Agent Skills Directory

[REMOTE_CODE_EXECUTION]: Provides instructions to download and run the official uv installation script from astral.sh. This targets a well-known service for Python tooling.
[COMMAND_EXECUTION]: Uses shell commands to synchronize dependencies (uv sync) and execute tests (uv run pytest) within the generated project directory to ensure functional correctness.
[DATA_EXFILTRATION]: No exfiltration patterns detected. The skill explicitly follows best practices by instructing the agent not to read the user's .env files and instead generate .env.example with placeholders for sensitive configuration.
[PROMPT_INJECTION]: No evidence of instructions to bypass safety guidelines or override agent behavior was found.
[EXTERNAL_DOWNLOADS]: Fetches project dependencies and environment management tools from official registries and the official astral.sh source.
[SAFE]: The skill incorporates strong security and operational best practices, including a mandatory confirmation step (hard gate) before performing write operations and clear warnings about schema verification when MCP discovery is unavailable.

kafka-python-client