robodog-terminal 0.2.2

Agentic coding terminal for the shell (runPixel gateway / OpenAI-compatible / offline backends)

robodog-terminal

A agentic agentic coding terminal: a prompted tool-use loop that reads/edits files, runs commands, runs tests, and self-corrects — over pluggable LLM backends. Designed to run leading models on self-hosted gateways (air-gapped, where an agentic coding terminal can't reach), and works equally with OpenAI-compatible models or a fully offline mock.

Install

pip install robodog-terminal          # or: pip install -e robodogcli/robodog

Run

robodog-terminal --backend openai --model gpt-4o     # live (OpenAI/OpenRouter)
robodog-terminal --backend gateway                       # enterprise box (the gateway / leading models)
robodog-terminal --echo                               # offline demo, no keys
robodog-terminal --backend openai -p "fix x.py and run the tests"   # headless (-p)
python -m terminal.run_tests                          # 18 test suites

Features

Agentic loop with an intent nudge + circuit breaker · tools (read/write/edit/ multi_edit/bash/run_script/run_tests/glob/grep/list_dir) with read-before-edit, post-edit syntax verification and fuzzy edits · foreground + background subagents (/bg /tasks /tail /kill) · per-prompt checkpoints with /rewind · JSONL sessions (/resume, --continue) · plan mode · skills & custom commands (.robodog/) · CLAUDE.md/ROBODOG.md hierarchy · a rich + prompt_toolkit TUI with an emoji/color status line, clickable file & file:line links, multiline paste, and mid-turn Ctrl+B backgrounding · headless -p (text/json) · /doctor.

Benchmarked at capability parity with a leading agentic coding assistant across 20 agentic scenarios. See robodogcli/docs/TERMINAL_MODE_PLAN.md for the full design.