The LLM Shortcut plugin streamlines prompt management by converting a folder of saved prompts into instantly accessible commands. Each file in the prompt directory becomes a command that can be triggered directly, with the current open note automatically passed as context to the chosen LLM provider. It supports any OpenAI -compatible API, giving flexibility in selecting your AI backend while keeping all activity local without external logging. The tree-like folder structure is preserved as a navigable list, making it easy to organize and access prompts for different use cases.
Obsidian plugins tagged with #openai-compatible. There are 2 plugins in this collection.
- Rank 1a year ago by Viktor Chernodub
- Rank 213 days ago by seanishoo
The Pick Me plugin annotates selected note text and PDF regions without modifying the source document. Each annotation is stored in a separate Markdown file linked back to its source. Note selections receive hidden anchor tags, while the built in PDF viewer supports rectangle captures that send extracted text and a screenshot to the configured model. Its highlighter snaps strokes to text lines and can merge adjacent highlights. An inline panel accepts templates or custom questions, preserves follow up history and reopens from an existing annotation. A vault wide index groups annotations by document, filters entries and separates stale anchors for rebinding. It can use any OpenAI compatible endpoint, keeps annotations and PDF rendering local and has no telemetry. The plugin is desktop only, and its API key is stored as plain text in the plugin settings file.