From 7df9dedfc986a9028f8fc59bc5ca211913a2529b Mon Sep 17 00:00:00 2001 From: Jon Saad-Falcon <41205309+jonsaadfalcon@users.noreply.github.com> Date: Sat, 14 Mar 2026 16:02:17 -0700 Subject: [PATCH] =?UTF-8?q?docs:=20refresh=20landing=20page=20=E2=80=94=20?= =?UTF-8?q?clearer=20primitives,=20research=20section,=20citation=20(#59)?= MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit - Rewrite Five Primitives with plain-English descriptions - Add 10+ engine backends with hyperlinks (Ollama, vLLM, SGLang, etc.) - Add Automated Workflows and Energy & Cost Tracking feature cards - Add Research section linking to Intelligence Per Watt and Stanford - Add Citation section with BibTeX - Fix hero-tagline max-width to span full title width Co-authored-by: Claude Opus 4.6 (1M context) --- docs/index.md | 43 +++++++++++++++++++++++++++----------- docs/stylesheets/extra.css | 2 +- 2 files changed, 32 insertions(+), 13 deletions(-) diff --git a/docs/index.md b/docs/index.md index a3b0ad54..b305f1a3 100644 --- a/docs/index.md +++ b/docs/index.md @@ -99,13 +99,15 @@ OpenJarvis is that stack. It is an opinionated framework for local-first persona --- -## Five Primitives +## Five Primitives for Personal AI -1. **Intelligence** — The LM: model catalog, generation defaults, quantization, preferred engine. -2. **Agents** — The agentic harness: system prompt, tools, context, retry and exit logic. Seven agent types. -3. **Tools** — MCP interface: web search, calculator, file I/O, code interpreter, retrieval, and any external MCP server. -4. **Engine** — The inference runtime: Ollama, vLLM, SGLang, llama.cpp, cloud APIs. Same `InferenceEngine` ABC. -5. **Learning** — Improvement loop: SFT weight updates, agent advisor, ICL updater. Trace-driven feedback. +OpenJarvis is built around five composable layers. Each has a clean interface and can be swapped independently. + +1. **Intelligence** — Pick a model, or let OpenJarvis pick one for your hardware. Manages the full catalog of local models across providers. +2. **Agents** — Multi-step reasoning with tool use. Seven built-in agent types from simple chat to orchestrated workflows. +3. **Tools** — Web search, calculator, file I/O, code interpreter, retrieval, and any external MCP server. +4. **Engine** — The inference runtime: [Ollama](https://ollama.com), [vLLM](https://github.com/vllm-project/vllm), [SGLang](https://github.com/sgl-project/sglang), [llama.cpp](https://github.com/ggerganov/llama.cpp), cloud APIs, and more. Auto-detects your hardware and recommends the best fit. +5. **Learning** — Your AI gets better over time. Every interaction generates traces that drive automatic improvements to model weights, prompts, and agent behavior. --- @@ -113,17 +115,17 @@ OpenJarvis is that stack. It is an opinionated framework for local-first persona
-- **Five Composable Primitives** +- **10+ Engine Backends** --- - Intelligence, Agents, Tools, Engine, and Learning — each with a clear ABC interface and decorator-based registry. + [Ollama](https://ollama.com), [vLLM](https://github.com/vllm-project/vllm), [SGLang](https://github.com/sgl-project/sglang), [llama.cpp](https://github.com/ggerganov/llama.cpp), [MLX](https://github.com/ml-explore/mlx), [Exo](https://github.com/exo-explore/exo), [LiteLLM](https://github.com/BerriAI/litellm), cloud (OpenAI/Anthropic/Google), and more. Same `InferenceEngine` interface, swap freely. -- **5 Engine Backends** +- **Automated Workflows** --- - Ollama, vLLM, SGLang, llama.cpp, and cloud (OpenAI/Anthropic/Google). Same `InferenceEngine` ABC. + Cron-based agents that monitor, summarize, and act. Code review, email triage, research digests — running 24/7 on your hardware. - **Hardware-Aware** @@ -143,11 +145,11 @@ OpenJarvis is that stack. It is an opinionated framework for local-first persona `jarvis serve` starts a FastAPI server with SSE streaming. Drop-in replacement for OpenAI clients. -- **Trace-Driven Learning** +- **Energy & Cost Tracking** --- - Every interaction is traced. The learning system improves models (SFT) and agents (prompt, tools, logic). + Built-in telemetry for GPU power draw, token costs, and latency. See exactly what each query costs in watts and dollars.
@@ -195,6 +197,23 @@ OpenJarvis is that stack. It is an opinionated framework for local-first persona +## Research + +OpenJarvis is part of [Intelligence Per Watt](https://www.intelligence-per-watt.ai/), a research initiative studying the efficiency of on-device AI systems. Developed at [Hazy Research](https://hazyresearch.stanford.edu/) and the [Scaling Intelligence Lab](https://scalingintelligence.stanford.edu/) at [Stanford SAIL](https://ai.stanford.edu/). + +Read the [blog post](https://scalingintelligence.stanford.edu/blogs/openjarvis/) for the full research motivation, architecture details, and experimental results. + +## Citation + +```bibtex +@misc{saadfalcon2026openjarvis, + title={OpenJarvis: Personal AI, On Personal Devices}, + author={Jon Saad-Falcon and Avanika Narayan and Herumb Shandilya and Hakki Orhun Akengin and Robby Manihani and Gabriel Bo and John Hennessy and Christopher R\'{e} and Azalia Mirhoseini}, + year={2026}, + howpublished={\url{https://scalingintelligence.stanford.edu/blogs/openjarvis/}}, +} +``` + ## Sponsors

diff --git a/docs/stylesheets/extra.css b/docs/stylesheets/extra.css index 7da9ea3f..1b56a47e 100644 --- a/docs/stylesheets/extra.css +++ b/docs/stylesheets/extra.css @@ -83,7 +83,7 @@ .hero-tagline { font-size: 1.05rem; color: var(--md-default-fg-color--light); - max-width: 600px; + max-width: 100%; line-height: 1.8; font-weight: 300; margin-bottom: 2rem;