The Agent Skills Directory

[COMMAND_EXECUTION]: The skill instructs the agent to use local shell commands, including cat and ls, to inspect project directory structures and configuration modules. It also uses python -c and pytest to validate that new configurations load correctly and that existing tests pass.
[PROMPT_INJECTION]: The skill has an indirect prompt injection surface because it ingests data from local project files (e.g., constants and configuration modules) using cat and then generates new code or modifications based on that input.
Ingestion points: Local source code files located at src/*/core/constants/_base.py, src/*/core/configs/_base.py, and src/*/core/configs/_main.py.
Boundary markers: No explicit markers are used to separate ingested file content from agent instructions.
Capability inventory: The agent can read files, write new Python modules to the filesystem, and execute Python code snippets via the CLI for validation.
Sanitization: There is no evidence of sanitization or filtering applied to the project files before the agent processes them.

configify