Chisel

unlisted

by XuYi

Process selected text with configurable AI actions.

Updated 23d agoMIT
View on GitHub

Chisel

Chisel is a lightweight Obsidian community plugin for selection-first AI actions. Select text in the editor, pick an action from the floating menu, and send the result to a popup, replacement, append, or insertion below the current paragraph.

Features

  • Floating action menu for selected editor text
  • Chinese and English UI switching
  • Immediate trigger or Cmd/Ctrl + Shift + A command trigger
  • Streaming responses with cancel support via Escape
  • Built-in actions: translate, polish, expand, summarize, explain, continue, proofread, questions, tags
  • Providers: OpenAI, Anthropic, Google Gemini, DeepSeek, and custom OpenAI-compatible endpoints
  • Provider model discovery through supported model-list endpoints, with manual model input as a fallback
  • Per-action visibility, ordering, output mode, and provider override
  • Custom actions with {{selection}}, {{language}}, {{filename}}, {{date}}, {{sourceLanguage}}, and {{targetLanguage}}
  • Custom action import/export as JSON
  • Multi-language translation target selection

Development

npm install
npm run build

For local Obsidian testing, copy or symlink this project folder into:

<vault>/.obsidian/plugins/chisel

Then enable the plugin in Obsidian community plugin settings.

Privacy

Chisel stores API keys locally through Obsidian's plugin data file. Selected text is sent only to the provider configured by the user for the action being run. Chisel does not collect telemetry.

Support

If Chisel saves you time, you can support ongoing maintenance on Ko-fi:

https://ko-fi.com/xuyi

Attribution

Provider logos are from Lobe Icons, licensed under MIT.

Release

Create a semantic version tag that matches manifest.json.

npm version patch
git push
git push --tags

The GitHub Actions release workflow builds the plugin and attaches main.js, manifest.json, and styles.css to the GitHub release.

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.