The Agent Skills Directory

[SAFE]: The skill implements a straightforward data processing pipeline. It reads structured input (CSV, YAML, JSONL) from a specified workspace and generates analytical reports and bindings in JSONL and Markdown formats. There is no network access, no execution of external untrusted code, and no use of risky functions like eval or exec. The Python script uses argparse for safe argument handling and relies on a local tooling module for utility functions, which is consistent with the agent environment's architecture. The file paths are constrained to the provided workspace directory.

section-bindings