Starstruk Logo MarkStarstruk

v3.6.0

DocumindHQ/documindv3.6.0Jul 18, 2026by assafelovic

AI Summary

This release focuses on stability and security, introducing deep agent integration with LangChain and fixing numerous robustness issues across retrievers, scrapers, and multi-agent systems.

Key Highlights

  • Integration with LangChain's deep agents framework (GPT Researcher as research engine)
  • Enhanced security measures including content sanitization and pinned dependencies (brotli)
  • Significant robustness improvements: 20 retriever guard fixes and 6 scraper robustness fixes
  • Real usage-based cost tracking and fixes for silently dropped token limits
  • Addition of Nebius Token Factory as a new LLM and embedding provider

New Features

  • Deep Agents example for LangChain
  • Nebius Token Factory provider support
  • Security.md documentation and untrusted content sanitization
  • Real usage-based cost tracking implementation
  • Organization of AG2 variant under multi_agents/ag2 directory

Full Release Notes

Introducing a super stable version with enhanced improvements across security, stability and performance. We're also introducing a collaboration with Langchain with the latest deep agents framework here - https://github.com/assafelovic/gpt-researcher/tree/main/deep_agents

Thank you and welcome to all the new contributors!

## What's Changed
* security: sanitize untrusted content, add SECURITY.md, pin brotli by @assafelovic in https://github.com/assafelovic/gpt-researcher/pull/1820
* Fix silently dropped token limits, real usage-based cost tracking, and research pipeline robustness by @assafelovic in https://github.com/assafelovic/gpt-researcher/pull/1861
* Add Deep Agents example: GPT Researcher as the research engine in a LangChain deep agent by @assafelovic in https://github.com/assafelovic/gpt-researcher/pull/1857
* Resolve post-merge conflicts: 4 retriever/context fixes (supersedes #1895-1898) by @assafelovic in https://github.com/assafelovic/gpt-researcher/pull/1906
* Retriever hardening: 20 guard fixes against malformed results (supersedes #1837…#1890) by @assafelovic in https://github.com/assafelovic/gpt-researcher/pull/1907
* Scraper robustness: 6 fixes (title/PDF-detection/temp-file/dimensions) (supersedes #1824…#1842) by @assafelovic in https://github.com/assafelovic/gpt-researcher/pull/1908
* Multi-agent robustness: bound revision loops + exact sentinels (supersedes #1883/#1885/#1886) by @assafelovic in https://github.com/assafelovic/gpt-researcher/pull/1909
* Core/misc hardening: 11 fixes (costs/query/llm/agent/mcp/config) (supersedes #1822…#1902) by @assafelovic in https://github.com/assafelovic/gpt-researcher/pull/1910
* Reconcile master→main: GetXAPI retriever, Claude 4.x temp fix, MiniMax M3, curate_sources fixes by @assafelovic in https://github.com/assafelovic/gpt-researcher/pull/1911
* feat: add Nebius Token Factory as LLM and embedding provider by @teionarr in https://github.com/assafelovic/gpt-researcher/pull/1891
* chore(multi_agents): nest AG2 variant under multi_agents/ag2 by @assafelovic in https://github.com/assafelovic/gpt-researcher/pull/1912
* docs(skill): fix drift in .claude reference docs (retrievers, multi-agents, api) by @assafelovic in https://github.com/assafelovic/gpt-researcher/pull/1913

## New Contributors
* @teionarr made their first contribution in https://github.com/assafelovic/gpt-researcher/pull/1891

**Full Changelog**: https://github.com/assafelovic/gpt-researcher/compare/v3.5.1...v3.6.0