The Agent Skills Directory

[PROMPT_INJECTION]: The skill directs the agent to automatically detect a project's Python version by reading local configuration files. This data ingestion creates a surface for indirect prompt injection, as the agent's behavior (specifically, which instruction files it loads) is influenced by the content of untrusted project files.
Ingestion points: The agent is instructed to read pyproject.toml, setup.py, setup.cfg, and .python-version files from the user's current project directory.
Boundary markers: The instructions do not specify any delimiters or safety markers to prevent the agent from being influenced by potentially malicious instructions embedded in these configuration files.
Capability inventory: The agent performs file-system read operations and conditionally loads additional Markdown instruction files based on the parsed version data.
Sanitization: No explicit sanitization or validation of the configuration file content is mentioned prior to the agent processing it for version detection.

dignified-python