PhrasePad

Local, privacy-first system-wide text expander & snippet manager for Windows 10/11 & macOS. Type a short trigger like ;addr or :sig anywhere — in your browser, email, editor, chat — and PhrasePad instantly expands it into full text, complete with dynamic tokens (date/time, clipboard), fill-in placeholders, and precise cursor placement.

Offline by default. No account, no cloud, no telemetry. Your snippets never leave your machine.

Overview

PhrasePad watches for abbreviation triggers you define and replaces them with expansions in real time, in any application. It combines a fast, always-available background expander with a friendly desktop manager for organizing snippets into groups, editing content, and testing expansions.

Core pieces:

• Background expander — a lightweight tray/menu-bar agent that detects triggers and performs the replacement via synthetic keystrokes.
• Snippet manager — a desktop GUI to create/edit/organize snippets, groups, and settings.
• phrasepad CLI — headless import/export, list, and expand-preview for scripting and backups.

Motivation

Everyone retypes the same things: email signatures, addresses, canned replies, boilerplate code, support macros, meeting notes headers. Cloud text-expanders exist, but they route your most personal snippets (addresses, phone numbers, replies) through third-party servers and require subscriptions. PhrasePad keeps everything local and free, works offline, and is cross-platform.

Use cases

• Email & support: ;greet, ;followup, ;refund → full canned replies with {clipboard} and {date} tokens.
• Developers: ;gpl, ;shebang, ;guid → license headers, boilerplate, generated GUIDs.
• Personal: ;addr, ;phone, ;iban → contact details typed once, reused everywhere.
• Forms/templates: ;invoice → a multi-field fill-in form that prompts for {name}, {amount}, {due} and inserts the assembled text.
• Writers: reusable phrases, footnotes, and formatting snippets with cursor left exactly where you want to keep typing.

How to use

Windows 10/11 quickstart

1. Download the latest phrasepad-win-x64.zip from Releases (or build from source).
2. Unzip and run PhrasePad.exe. It starts in the system tray.
3. Open the manager (double-click tray icon), click New Snippet, set a trigger (;addr) and expansion text.
4. Switch to any app and type ;addr — it expands instantly.

macOS quickstart

1. Download PhrasePad.dmg from Releases (or build from source).
2. Drag PhrasePad.app to Applications and launch it (menu-bar app).
3. Grant Accessibility permission when prompted (System Settings → Privacy & Security → Accessibility) — required to detect typing and send keystrokes.
4. Add a snippet in the manager, then type your trigger in any app.

Example workflow

Trigger:   ;sig
Expansion: Best regards,
           Alex Rivera
           {date:yyyy-MM-dd}

Typing ;sig in your mail client produces:

Best regards,
Alex Rivera
2026-08-15

Placeholder form example:

Trigger:   ;mtg
Expansion: Meeting notes — {topic}
           Attendees: {who}
           Action items:
           - {cursor}

Typing ;mtg pops a small fill-in dialog for topic and who, then inserts the text with the cursor parked on the action-items line.

CLI

phrasepad list                       # list all snippets
phrasepad list --json                # machine-readable library
phrasepad expand ";sig"              # preview with an ⟦cursor⟧ marker
phrasepad expand ";invoice" \
  --field name=Ada --field amount=42 # fill placeholders without typing
phrasepad export snippets.json       # back up your library
phrasepad import snippets.json       # merge without replacing conflicts

The CLI uses the same platform library as the GUI by default. Pass --library <path> to work with a specific library file. list, expand, import, and export accept --json where structured output is useful. Import conflicts leave the existing snippet or group unchanged and return exit code 4; usage errors return 2, missing triggers return 3, and storage or JSON errors return 1.

Local-AI integration (optional, off by default)

PhrasePad can optionally call a local small model to help you author snippets — never to run your expansions. When enabled and reachable, it talks to an OpenAI-compatible endpoint on localhost (e.g. Ollama or llama.cpp) using small models such as Llama 3.2, Qwen2.5, Phi-3-mini, or MiniCPM class.

• Rewrite: tighten/shorten/adjust tone of a snippet you're editing.
• Generate: describe a canned reply and get a draft snippet to save.
• Suggest triggers: propose a short abbreviation for new snippets.

All AI features are opt-in, run locally, include a reachability probe, and degrade gracefully to manual editing if no model is available. Expansion at type-time is always deterministic and never depends on AI.

Current status / milestones

Early scaffold. Tracking issues drive incremental delivery:

• M1 — Core snippet model + storage + trigger matching engine
• M2 — Windows background expander (hook + keystroke injection)
• M3 — macOS background expander (Accessibility + CGEvent)
• M4 — Snippet manager desktop UI (Avalonia)
• M5 — Dynamic tokens + placeholder fill-in forms
• M6 — phrasepad CLI (import/export/list/expand)
• M7 — Optional local-AI authoring assist
• M8 — Packaging & CI (Windows zip/MSIX, macOS .app/.dmg)

See PLAN.md for scope, architecture, and non-goals.