allenhutchison/obsidian-gemini

An obsidian plugin to interact with Google Gemini

Gemini Scribe – AI‑powered assistant inside Obsidian

What it is

  • An Obsidian community‑plugin that lets you talk to large‑language‑model (LLM) providers (Google Gemini, Ollama, OpenAI, Anthropic) from within your vault. The plugin treats your notes as context, so the model can read, create, edit, and organise files for you.

Core capabilities

  • Agent mode with tool‑calling – the model can run built‑in tools to search the vault, read/write notes, move/rename files, create folders, and perform deep research with citations. Sessions are saved as markdown files, so you can version‑control them.
  • Per‑feature provider routing – you decide which provider (Gemini, Ollama, OpenAI, Anthropic) handles each task: chat, summarisation, completions, rewrite, web search, RAG, image generation, etc.
  • Semantic vault search – uses Google’s File Search API to index your notes (including PDFs and attachments) and retrieve results by meaning rather than keyword.
  • Smart summarisation & rewrite – one‑click commands to summarise a note into its front‑matter, rewrite selected text, or ask a question about a selection.
  • IDE‑style completions – real‑time, context‑aware suggestions as you type, accepted with Tab.
  • Plan mode – optional UI that shows a step‑by‑step plan before the agent executes actions.
  • Scheduled tasks & lifecycle hooks – automate recurring prompts or trigger head‑less agent runs when files are created/modified/etc.
  • Projects & Skills – group settings, prompts, and permissions per folder‑based “project”; load extensible skill packages (agentskills.io spec) to give the agent specialised workflows.
  • Image paste & generation – paste images into the chat, or generate images via Gemini/OpenAI models when the provider supports it.
  • MCP (Model Context Protocol) support – experimental integration with external tool servers via stdio or HTTP, with OAuth and per‑tool trust settings.

How to get started

  1. Install Gemini Scribe from Obsidian’s Community Plugins list.
  2. Obtain an API key (Google AI Studio for Gemini, or your own key for OpenAI/Anthropic; Ollama needs no key).
  3. Open Settings → Providers and add the desired provider(s).
  4. On the Features page, route each feature to the provider you prefer.
  5. Launch the agent chat via the ribbon icon or the command palette (e.g., “Gemini Scribe: Open agent chat”).
  6. Use the right‑click context‑menu commands for selected‑text actions, or start a scheduled task from the Automation settings.

Supported environments

  • Desktop Obsidian (Windows, macOS, Linux) – full feature set.
  • Mobile Obsidian – UI fixes allow the agent view and most commands to work on iOS/Android.
  • Local models via Ollama or any OpenAI‑compatible server (LM Studio, MLX, etc.) for offline use.

Recent highlights (v4.12.1)

  • Passed Obsidian’s plugin audit (no suppressed checks).
  • Mobile‑friendly agent view.
  • Conditional registration of the image‑generation tool based on provider connectivity.\n- Added Anthropic and generic OpenAI‑compatible providers, with streaming and tool‑calling support.
  • Redesigned settings UI and per‑feature routing table.
  • Improved robustness of tool‑call handling, loop detection, and token usage display.

Who might benefit

  • Knowledge‑workers who keep large note collections and want AI to organise, summarise, or extend them.
  • Researchers needing automated deep‑research runs with citations directly in their vault.
  • Writers who like AI‑assisted drafting, rewriting, and idea generation without leaving Obsidian.
  • Teams that store notes in a git‑backed vault and want reproducible AI sessions saved as markdown.

Where to learn more

All details above are taken directly from the repository’s README; no additional features have been inferred.

Related

  • Project
    nhaouari/obsidian-textgenerator-plugin

    An open-source AI assistant plugin for Obsidian that enables users to generate text, summaries, and ideas directly within their knowledge base using various LLM providers.

  • Project
    Lapis0x0/obsidian-yolo

    YOLO is an Obsidian community plugin that embeds AI agents directly into the note‑taking app. It provides chat, multi‑model support, retrieval‑augmented generation from your vault, tool‑calling, a learning mode with flashcards, multi‑window chats, memory, and a fast AI‑whiteboard. Install via the plugin store, configure an API key (OpenAI, Claude, Gemini, etc.), and start using Quick Ask (`@`) or the sidebar. The project is MIT‑licensed and actively maintained.

  • Project
    yaroslavyaroslav/OpenAI-sublime-text

    The repository provides a genuine Sublime Text plugin that integrates large‑language‑model APIs (OpenAI, Anthropic, Gemini, and any OpenAI‑compatible service) into the editor, offering chat, inline phantom responses, project‑specific histories, and flexible context handling.

  • Project
    algorithmicsuperintelligence/optillm

    OptiLLM is an OpenAI‑compatible inference proxy that applies 20+ reasoning‑boost techniques (best‑of‑N, chain‑of‑thought, MCTS, Mixture‑of‑Agents, etc.) at request time, giving 2‑10× higher accuracy on math, coding and logic tasks without any model training. It works with any OpenAI‑compatible provider via LiteLLM, offers a plugin system (memory, privacy, URL fetch, code execution, web search), and can be installed via pip or Docker.

  • Project
    hkcanan/katmer-code

    An Obsidian plugin that integrates multiple AI providers (Claude, Gemini, Codex, Antigravity) into a sidebar, featuring inline diff editing and specialized academic research tools.