Laibrarian

pending

by Matthias

Summarize selected notes or the active note with LLM prompt presets and save generated markdown outputs.

Updated 1mo agoMITDiscovered via Obsidian Unofficial Plugins
View on GitHub

laibrarian

Generate markdown summaries from selected Obsidian notes using multiple configurable AI providers.

Features

  • Select any number of markdown files from a vault-wide picker modal.
  • Send selected notes to an LLM and save the output into a configured folder.
  • Send only the currently open note with a selected preset and create a sibling output file.
  • Provider selector with an 8-provider set:
    • OpenAI
    • Anthropic
    • Google Gemini
    • Google Cloud Vertex AI
    • Azure OpenAI
    • ElevenLabs (provider slot included for parity; not yet implemented as direct chat-completions backend)
    • AWS Bedrock (provider slot uses the legacy internal ID aws-polly for compatibility)
    • OpenAI Compatible
  • Built-in prompt presets (editable in settings):
    • TTS fluid summary
    • Bullet summary
    • Interesting links
    • Exploration ideas

Commands

  • Summarize selected markdown notes
  • Summarize current note with prompt preset

Provider settings (summary)

  • OpenAI: API key + model
  • Anthropic: API key + model
  • Gemini (AI Studio): API key + model (endpoint preconfigured)
  • Google Cloud Vertex AI: OAuth access token + project ID + location + model (endpoint auto-built; optional override)
  • Azure OpenAI: API key + resource endpoint + deployment ID + API version (full endpoint also supported)
  • AWS Bedrock: API key + region + model (endpoint auto-built; optional override)
  • ElevenLabs: docs links and guidance only for now (no direct summarization backend yet)
  • OpenAI Compatible: API key (optional for localhost) + base URL + model

Output behavior

  • Multi-file command writes to Output folder for vault summaries.
  • Active-file command writes next to the source note as:
    • <original basename> - <preset suffix>.md

Community submission disclosures

  • This plugin sends note content to the selected third-party AI provider over the network.
  • Using this plugin generally requires provider accounts and API keys, and provider billing may apply.
  • API keys are stored locally in Obsidian plugin data (data.json). Do not commit that file.
  • The plugin includes no telemetry, ads, or self-update mechanism.

Acknowledgements and inspirations

This plugin takes inspiration from community work that is directly relevant to multi-provider summarization workflows:

  • ai-providers / local-gpt by Pavel Frankov for multi-provider settings architecture.
  • scribe by Mike Alicea for structured transformation/summarization workflow framing.

Credit where due:

  • Matthias provided the product vision, selected scope, and performed iterative testing/validation.
  • Codex implemented most of the code and release preparation steps under Matthias's direction.

Build

npm ci
npm run build

Release

  • GitHub Actions release workflow: .github/workflows/release.yml
  • Required release assets: main.js, manifest.json, and styles.css
  • Git tag must match manifest.json.version exactly (no v prefix)

For plugin developers

Search results and similarity scores are powered by semantic analysis of your plugin's README. If your plugin isn't appearing for searches you'd expect, try updating your README to clearly describe your plugin's purpose, features, and use cases.