page2md

convert and doc_to_md both fetch a URL and return Markdown, with doc_to_md adding PDF and chunking support; the boundary is only partly clarified by descriptions. convert_raw is clearly distinct (raw HTML input) and trial_key is unrelated to conversion, but the two fetch-and-convert tools risk misselection.

All names use snake_case, but conventions vary: convert is a bare verb, convert_raw is verb_modifier, doc_to_md is noun_to_format, and trial_key is noun_noun. The set is readable but does not follow a single predictable verb_noun pattern.

Four tools is well within the appropriate range for a focused page-to-Markdown conversion service. Each tool has a plausible role, though convert and doc_to_md overlap enough that one could be considered redundant.

The surface covers the main lifecycle: fetch-and-convert, convert existing HTML, PDF/chunked output for RAG, and key acquisition. Minor gaps exist around configuration options (e.g., chunk size, output formatting) and batch conversion, but core workflows are covered.