# tiny-context MCP server

> Enables AI agents to query, search, and extract data from local files like PDFs, spreadsheets, and logs without reading entire files into context.

- Repository: https://github.com/Warddamn/tiny-tools
- Page: https://hostedmcp.org/servers/warddamn/tiny-context
- Transport: stdio
- Stars: 0
- License: MIT
- Latest version: 0.3.0

## Tools (10)

- `query_table`
- `summarize_log`
- `file_map`
- `query_file`
- `read_section`
- `extract`
- `diff_files`
- `validate_file`
- `zod`
- `commander`

## FAQ

### What file formats does tiny-context support?

The tools support PDF, DOCX, PPTX, XLSX, CSV, Parquet, TXT, Markdown, and JSON/JSONL/NDJSON.

### Is my data safe when using these tools?

The tools process local files and do not upload them or call a model themselves; they also include safety controls like read-only SQL queries and disabled external network access.

### Does the server send telemetry or usage reports?

No, there is no telemetry, usage reports, or analytics sent by the server.

### Can I reuse extraction settings for similar documents?

Yes, you can save document profiles to extract data from recurring forms or reports in batches.