Summary
Somewhere to put the twenty documents you currently keep grepping through.
The whole surface is add, list, search, inspect, remove, clear, which is enough for a working set you assemble for one project and throw away afterwards. Updating a document means removing it and adding it again, and anything outside the supported extensions never enters the index at all.
What it is
A semantic knowledge base in six tools — documents in, vector search out — over local files, GitHub repositories and direct file URLs.
What you get
- add_document accepts a local file, a GitHub repository or a direct file URL
- Code and prose in the same index: .py, .js, .ts, .tsx, .java, .cpp, .go, .rs, .rb and .php alongside .md, .txt, .json, .yaml, .toml and shell scripts
- search_documents matches on AI embeddings, so a query for authentication reaches text about login and credentials without those words appearing
- The index stays inspectable and reversible: list_documents pages through what is stored, get_document_info reports one document's statistics, and removal works per document or all at once
Requirements
No account and no key. Documents are capped at 10 MB each, raised through the MAX_FILE_SIZE_MB environment variable.
Setup effort
One command — uvx context-lens