feat: add tika_parser

[amitjaju82]

Tika Document Parser

Universal document parsing pipeline via Apache Tika. Handles PDF, Office, images (automatic OCR), email, archives, HTML, and 1000+ formats with a single tool call.

Features

• Universal format support - PDF, Office, images (OCR), email, archives, HTML, XML, RTF, and 1000+ formats
• Automatic OCR - Scanned PDFs and standalone images are OCR'd via Tika's built-in vision model
• Recursive extraction - Containers (ZIP, DOCX with attachments, email with attachments) are fully extracted with per-object provenance
• Production-ready - Async client with concurrency cap, retries with exponential backoff, SHA-256 dedup cache, and size guard
• Agent tool - tika_parse tool with parse, parse_url, batch, and health actions
• Settings UI - Full configuration panel in Settings > Agent

Links

• GitHub: https://github.com/amitjaju82/tika_parser
• Author: Amit Jaju (amitjaju@gmail.com)
• Tags: tools, integration, workflow