The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes shell commands to analyze the repository structure and run the pytest framework.
Evidence: python scripts/analyze_repo_structure.py <repo_path> and various pytest commands are documented in the workflow.
[REMOTE_CODE_EXECUTION]: The skill implements new Python functionality and executes it at runtime to verify its correctness.
Evidence: The workflow explicitly instructs the agent to implement a feature and then run pytest to execute that code.
[PROMPT_INJECTION]: The skill processes untrusted natural language descriptions and uses them to generate code that is subsequently executed, creating a surface for indirect prompt injection.
Ingestion points: Natural language feature descriptions and the content of files in the repository.
Boundary markers: No delimiters or specific safety instructions are used to distinguish the feature request from the system prompts.
Capability inventory: The skill has the ability to write Python files and execute them through a shell.
Sanitization: No sanitization or validation of the generated code is described before the testing phase begins.

incremental-python-programmer