What problem does Agent Terminal solve?

It bridges voice input, semantic memory, and AI task delegation into one desktop terminal so you can speak a goal and have it routed to the right tool without leaving your workflow.

Roadmap

The roadmap prioritizes reliable daily usage for solo developers before any shared-session or team-operation story.

Shipped in the current app

Tauri native desktop shell (Linux/macOS/Windows)
5-tab UI: Composer, Runs, History, Transcripts, Context
Pre-send guard (pure, no side effects)
SQLite session history at ~/.agent-terminal/history.db
Agent Brain memory panel in every session
Agent Harness task launch (POST /terminal/launch)
Wispr ingestion endpoint (POST /api/wispr/ingest)
Live Harness run status in Runs tab

Current focus — operator feedback

The next work is focused on making the local loop easier to understand and operate:

Richer Brain recall at session start
Keyboard-first shortcuts for power users
Improved session search in History tab
More transparent guard, context, and Harness execution feedback

Much later — Team-ready operation

Shared session history, multi-user Brain workspaces, and remote Harness controller pairing.

◆Public roadmap boundary

Standalone local use remains the primary product path. Private prompts, custom evaluation designs, internal thresholds, and unreleased deployment details are intentionally not part of this public roadmap.