0.9.1

Adds support for internationalization, now allowing you to specify the language for knowledge assets, the Ask Q&A mode, and the GUI.

Full Changelog: 0.9.0-stable...0.9.1

0.9.0

Terrain v0.9.0

Terrain prepares the ground so AI agents don't have to guess where to stand.

An engineering environment manager for humans and AI coding assistants — knowledge as the map, tools as the roads, conventions as the trail markers. This first stable release ships the full platform built on the Litho (deepwiki-rs) documentation engine.

Highlights

• Knowledge as the map — auto-generated, always-in-sync C4 docs and agent context, produced from your code and consumed by both humans and agents
• Dual-track docs — narrative human/ docs for people, structured agent/context.md + grep-friendly repomix.md for agents, with freshness scoring and incremental updates
• AI-native — speaks ACP, so Claude Code, Codex, OpenCode, and Cursor read the same knowledge layers through terrain tools
• Tools as the roads — terrain env apply installs Skills, CodeGraph, RTK, and AGENTS.md in one command
• Conventions as trail markers — four-phase SDD workflow (requirements → design → codegen → review) producing reviewable Markdown artifacts
• Desktop app + CLI — browse, search, and ask questions over your codebase with the integrated Litho Book reader and DeepWiki-style Q&A

Links

• README
• License: MIT