A DeepSeek-native AI coding agent for the terminal that keeps token costs low across long sessions through prefix-cache stability. It is distributed as a single static Go binary and configured through a TOML file.
Project overview
A DeepSeek-native AI coding agent for the terminal that keeps token costs low across long sessions through prefix-cache stability. It is distributed as a single static Go binary and configured through a TOML file.
Project type
MCP · AI Agent · AI Coding
Use cases
Coding & Development · Automation
Deployment
Refer to project documentation
License
MIT
Best for
Developers and AI engineers who work in the terminal and want to maintain low token costs across long coding sessions using DeepSeek models.
Teams requiring a self-hosted, local coding agent with external tool integrations routed through MCP-compatible subprocesses.
Key capabilities
Providers, the agent, enabled tools, and plugins are declared in reasonix.toml with no hardcoded models.
DeepSeek ships as a preset; any OpenAI-compatible endpoint is a config entry. Optionally run two models together (executor + planner) in separate, cache-stable sessions.
External tools run as subprocesses over stdio JSON-RPC (MCP-compatible). Built-in tools self-register at compile time.
Startup injects a small stable environment summary, stale tool output is snipped/pruned before summary compaction, and the built-in tool schema contract is documented for regression review.
CGO_ENABLED=0 single binary; cross-compile to six targets with one command. The only dependency is a TOML parser.
Supports slash commands and @ references for interaction.
Create, share, preview, run, edit, and safely delete isolated agent profiles from desktop or CLI.
Snapshot-based edit safety net (Esc-Esc / /rewind).
Limitations and risks
The system is engineered specifically around DeepSeek's prefix cache stability, though other models can be configured.
Operation requires external API access, such as the DeepSeek API, which may incur paid service costs.
Getting started
Install via npm or brew with a single binary. Configure a DeepSeek API key through setup or environment variables, create a minimal reasonix.toml file, and run reasonix to start the agent.
Evidence and sources
GitHub project description: DeepSeek-native AI coding agent for your terminal. Engineered around prefix-cache stability — leave it running.
README: A config- and plugin-driven harness — a single static Go binary, tuned around DeepSeek's prefix cache so token costs stay low across long sessions.
README: A DeepSeek-native AI coding agent for your terminal.
README: Zero-friction distribution. CGO_ENABLED=0 single binary; cross-compile to six targets with one command. The only dependency is a TOML parser.
README: Windows builds are code-signed with a free certificate provided by the SignPath Foundation, with signing through SignPath.io.