🔥 sponsors/khoj-ai
Your AI second brain. Self-hostable. Get answers from the web or your docs. Build custom agents, schedule automations, do deep research. Turn any online or local LLM into your personal, autonomous AI (gpt, claude, gemini, llama, qwen, mistral). Get started - free. — Trending on GitHub today with 48 new stars.
TLDR; A trustworthy, personal assistant accessible to everyone at their fingertips. That's the vision for Khoj 🏮
Hi 👋🏽!
We're Debanjum and Saba. We started Khoj to get better, deeper assistance in our day to day lives. We're both avid note takers, but we found that the tools to use our personal knowledge bases lagged far behind what was possible.
While computing capabilities have advanced at a staggering rate over the last few decades, our individual capabilities have stagnated*. We don't think that there's any reason to accept that as a necessary byproduct.*
Khoj already allows you to search and chat with your documents and images on consumer hardware. Get blazing fast, natural language search that works completely offline. Get answers and build on top of your existing knowledge base using Chat when connected to the internet.
There are massive improvements we need to make before this is ready for all. We're planning to add:
-
A desktop client with cloud syncing, so you can access your assistant wherever you are.
-
Support for open source LLMs, so you don't have to sacrifice privacy for assistance.
-
A mobile client that allows everyday users to have deeper assistance in their day-to-day lives.
We've decided to work on Khoj full time because we believe we're at a critical juncture to shape our engagements with AI. Sponsorship provides the vital social support we need to fully realize our long term vision for Khoj ✨
Sign in to highlight and annotate this article

Conversation starters
Daily AI Digest
Get the top 5 AI stories delivered to your inbox every morning.
More about
githubtrendingopen-sourceThe Rogue Replication Threat Model
An illustration of a sequence of events where rogue replicating agents emerge and cause harm. In 2023, METR 1 introduced the term “Autonomous Replication and Adaptation” (“ARA”) , which refers to the cluster of capabilities required for LLM agents to acquire and manage resources, evade shutdown, and adapt to novel challenges. Since then, the autonomous replication concern has become more mainstream. At the AI Seoul Summit , 27 nations agreed on thresholds where “model capabilities could pose ‘severe risks’ without appropriate mitigations,” including “autonomous replication and adaptation.” OpenAI , Anthropic , and Google DeepMind have also included autonomous replication evaluations in their safety frameworks. This blog post presents some of our thoughts on the most commonly recognized var
1.13.0a3
<h2>What's Changed</h2> <h3>Features</h3> <ul> <li>Emit token usage data in LLMCallCompletedEvent</li> <li>Extract and publish tool metadata to AMP</li> </ul> <h3>Bug Fixes</h3> <ul> <li>Handle GPT-5.x models not supporting the <code>stop</code> API parameter</li> </ul> <h3>Documentation</h3> <ul> <li>Fix inaccuracies in agent-capabilities across all languages</li> <li>Add Agent Capabilities overview and improve Skills documentation</li> <li>Add comprehensive SSO configuration guide</li> <li>Update changelog and version for v1.13.0rc1</li> </ul> <h3>Refactoring</h3> <ul> <li>Convert Flow to Pydantic BaseModel</li> <li>Convert LLM classes to Pydantic BaseModel</li> <li>Replace InstanceOf[T] with plain type annotations</li> <li>Remove unused methods</li> </ul> <h2>Contributors</h2> <p><a cla
b8598
<details open=""> <p>sync : ggml</p> </details> <p><strong>macOS/iOS:</strong></p> <ul> <li><a href="https://github.com/ggml-org/llama.cpp/releases/download/b8598/llama-b8598-bin-macos-arm64.tar.gz">macOS Apple Silicon (arm64)</a></li> <li><a href="https://github.com/ggml-org/llama.cpp/releases/download/b8598/llama-b8598-bin-macos-x64.tar.gz">macOS Intel (x64)</a></li> <li><a href="https://github.com/ggml-org/llama.cpp/releases/download/b8598/llama-b8598-xcframework.zip">iOS XCFramework</a></li> </ul> <p><strong>Linux:</strong></p> <ul> <li><a href="https://github.com/ggml-org/llama.cpp/releases/download/b8598/llama-b8598-bin-ubuntu-x64.tar.gz">Ubuntu x64 (CPU)</a></li> <li><a href="https://github.com/ggml-org/llama.cpp/releases/download/b8598/llama-b8598-bin-ubuntu-vulkan-x64.tar.gz">Ubun
Knowledge Map
Connected Articles — Knowledge Graph
This article is connected to other articles through shared AI topics and tags.
More in Open Source AI
b8598
<details open=""> <p>sync : ggml</p> </details> <p><strong>macOS/iOS:</strong></p> <ul> <li><a href="https://github.com/ggml-org/llama.cpp/releases/download/b8598/llama-b8598-bin-macos-arm64.tar.gz">macOS Apple Silicon (arm64)</a></li> <li><a href="https://github.com/ggml-org/llama.cpp/releases/download/b8598/llama-b8598-bin-macos-x64.tar.gz">macOS Intel (x64)</a></li> <li><a href="https://github.com/ggml-org/llama.cpp/releases/download/b8598/llama-b8598-xcframework.zip">iOS XCFramework</a></li> </ul> <p><strong>Linux:</strong></p> <ul> <li><a href="https://github.com/ggml-org/llama.cpp/releases/download/b8598/llama-b8598-bin-ubuntu-x64.tar.gz">Ubuntu x64 (CPU)</a></li> <li><a href="https://github.com/ggml-org/llama.cpp/releases/download/b8598/llama-b8598-bin-ubuntu-vulkan-x64.tar.gz">Ubun
The Claude Code Leak Proved What We've Been Building For
<p>Today Anthropic accidentally shipped 512,000 lines of Claude Code's source code to npm. A source map file that should have been stripped from the build made it into version 2.1.88 of the @anthropic-ai/claude-code package. Within hours, the entire codebase was mirrored on GitHub and dissected by thousands of developers.</p> <p>The leak itself was a packaging error. Human mistake. It happens.</p> <p>But what the leak <em>revealed</em> is the part that matters.</p> <h3> The Real Problem Isn't the Leak </h3> <p>Check Point Research had already disclosed CVE-2025-59536 back in October — a vulnerability where malicious <code>.mcp.json</code> files in a repository could execute arbitrary shell commands the moment you open Claude Code. No trust prompt. No confirmation dialog. The MCP server ini
Meteorite found in Somalia turns out to contain two minerals that are not found on Earth - Interesting Engineering
<a href="https://news.google.com/rss/articles/CBMiiwFBVV95cUxNb01VeTJqS2NyZjNzdnNLYTFBTW9JUEJoZkV2ekUzNjlMZ2hYemRiMWw5dXBTUUY4cFpNbWg5U0trLUVheHkzeWFXb1dsNnhULV94SDlFOElFRFhsU3hORVFJSnRaYU9KdUd2Q2JmckdVNWJhWVJveWZEVkQxd19SWjgxSjd2TFFDUFZr?oc=5" target="_blank">Meteorite found in Somalia turns out to contain two minerals that are not found on Earth</a> <font color="#6f6f6f">Interesting Engineering</font>
Erdogan holds high-level talks with Sudan's leader in Ankara - Yeni Safak English
<a href="https://news.google.com/rss/articles/CBMiogFBVV95cUxNVVlVMWM0U1lBNVJ4dkJzLS1Nc3k1RDFya3hseERiSm5mVC1LaGN4NjJhNnkybjB6Sk94WDcteXd5MWRYTTNXN1VZejI2bnpWYjVVUE91VUlTSTlsc2JvYkxCd1NUVXdjOEhndzhTRllKaENhNWlhb0tTTkdDSXZzU0E3RC1FelNtMWszVVlGc1R0NEVtYk9QX1VzMnpPbzNwZHc?oc=5" target="_blank">Erdogan holds high-level talks with Sudan's leader in Ankara</a> <font color="#6f6f6f">Yeni Safak English</font>
Discussion
Sign in to join the discussion
No comments yet — be the first to share your thoughts!