Back to directory

DSCODE

DSCODE is a terminal coding agent for macOS, built on DeepSeek Harness, in which multiple agent sessions on the same machine can see each other, hand off tasks, review one another's diffs, and have their approvals decided by an independent reviewer rather than a fixed rule table.

DSCODE

DSCODE runs a persistent shell that reads and writes code and runs tests, shared by its terminal UI, its command-line interface and any scripts, so all of them operate inside one session runtime instead of each starting its own. It installs as a pinned, versioned harness, with its DeepSeek Harness dependencies, terminal UI and plugins verified together.

Sessions on the same machine are visible to each other: a side question can run in its own read-only child session and answer in a panel without entering the main conversation; a task can be handed to another session, which can read the requesting session's transcript, answer, and hand the result back; and tool-call approvals are decided by an independent reviewer acting on the user's own instruction rather than a static rule table, with a record of what it allowed and what it cost. A persisted mailbox, retry deduplication and a finite budget bound how many messages, duplicate runs or wake-up loops can occur between sessions.

It is written in JavaScript for Node.js and runs on macOS.

More in Coding agents

See all 793
  • The City

    github.com

    A native macOS app that shows Claude Code runs as a city, with each project as a building, each team as a floor of robots, and every animation driven by a real event from the claude CLI's stream.

  • rig

    tryrig.ai

    A small, single-binary agent runtime that points at any OpenAI-compatible model, local or hosted, reads and edits a repository's code, runs its tests and remembers what it learned, with a loop implemented in a few hundred lines.

  • Jauvex

    jauvex.reindent.com

    An Electron desktop app that puts Claude Code, Codex and Grok coding-agent sessions side by side and lets you steer them by voice, with a small model handling fast routing decisions.