Summary
The exclusions come from the repository's own .gitignore, so the dump matches what the project already treats as source.
Binary files are skipped and ignored paths stay out without a second exclusion list to keep in step with the first. Chunking is the caller's decision rather than the tool's: num_chunks and chunk_index return the repository in pieces you ask for in order, and because each piece keeps its per-file path headers, a chunk still reads as itself rather than as a fragment.
What it is
A single tool that walks a directory and returns its text files as one concatenated document, each file wrapped in headers and footers carrying its path.
What you get
- dump_codebase_context, which recursively reads the text files under a base_path, respecting .gitignore rules and skipping binary files.
- File path headers and footers around each file's content, so the boundaries survive the concatenation.
- Chunking through num_chunks and chunk_index, for a codebase too large to come back in one response.
Requirements
Nothing to supply — no account, no key. A directory the server can read, and a .gitignore in it if you want the exclusions the project already maintains to apply.
Setup effort
One command — npx -y @lex-tools/codebase-context-dumper