Skip to main content
prompt GP ChatGPT

hybrid · open-rua, behavior — Claude Code prompt for R1Pro control

The hybrid prompt for Claude Code directs R1Pro through actor_act. It covers bounded local phases using the robot’s whole body and grippers.

MIT ★ 488 updated 2026-11-15

Install into Claude Code · Manually: put the files in place All methods and platforms ↓
.claude/commands/hybrid.md
Open on GitHub

What hybrid does

The prompt uses actor_act with the whole_body group for bounded local phases on R1Pro. That group couples the mobile base, torso, arms, and both grippers. Partial body groups and direct gripper control are unavailable in this profile.

The actor returns 32 candidate samples; the default execution prefix is 16 samples at 30 Hz. Each continuation takes a fresh observation. Simulation advances only while acknowledged commands run, then pauses while the agent or actor reasons. Previews show requested commands, not predicted collisions or object motion.

Inspect fresh RGB observations before marking a local phase complete. Tools do not report official task predicates or success; the native evaluator records the outcome after finish. Completing an execution request does not establish task success.

What it helps with

  • control the base, torso, arms, and grippers as one group
  • set base motion in the robot-local frame
  • inspect RGB observations before ending a phase
  • separate command execution from task success

Example prompts for the agent

  • Move R1Pro through this local phase using its whole body.
  • Move the base forward and set the arm targets for a grasp.
  • Inspect fresh RGB observations before ending the local phase.

How to install hybrid

Manually: put the files in place

.claude/commands/hybrid.md

Questions about hybrid

How do I install hybrid in Claude Code?

Add the command file at .claude/commands/hybrid.md.

What does hybrid do?

It directs R1Pro through actor_act, covering the base, torso, arms, and both grippers. The agent gets a fresh observation after each continuation.

Does hybrid confirm that a task succeeded?

No. Tools do not return official success predicates; the native evaluator records the outcome after finish.

Source text

hybrid.md · MIT
Use actor_act with active_groups=["whole_body"] for bounded local phases.
R1Pro couples its mobile base, torso, arms and both grippers. Partial body groups
and Direct gripper tools are unavailable in this profile.

The actor returns 32 candidate samples; the default execution prefix is 16 samples
at 30 Hz. Every continuation observes again. Simulation advances only while
executing acknowledged commands and pauses while the agent or actor reasons.

Base commands use the robot-local frame: forward/lateral metres per second and
yaw radians per second. Arm and torso commands are absolute joint targets in
radians; the two fingers of each gripper receive symmetric metre targets.
Raw actor gripper values are normalized commands, not measured opening widths.

Previews describe requested commands, not predicted collisions or object motion.
The optional reviewer may approve a shorter actor prefix or stop; it cannot edit
individual joints, move a gripper directly, or guarantee a stable grasp.
Inspect fresh RGB observations before declaring a local phase complete.

Official task predicates and success are unavailable as tool feedback. The native
evaluator records the outcome after finish. This profile retains native assisted
grasping; completion of an execution request is not evidence of task success.

© Open-RUA · MIT · Open-RUA/uni-robot-agent

Similar prompts

All prompts for AI and ML →

hybrid by other authors comes first.

hybrid Open-RUA The hybrid Claude Code prompt coordinates a humanoid as one actor. It combines upper-body targets with walking velocity while native control handles the legs and waist. GP ChatGPT ★ 488repo stars hybrid Open-RUA The hybrid prompt for Claude Code reports RoboDojo task progress with evidence-backed subgoals. A reviewer checks action-model proposals before execution. GP ChatGPT ★ 488repo stars hybrid Open-RUA The hybrid Claude Code prompt controls an arm and dexterous hand assembly in DexJoCo. It supports pick_bucket with the published single-arm model recipe. GP ChatGPT ★ 488repo stars hybrid Open-RUA The hybrid Claude Code prompt helps choose between actor_act and direct robot motion. It guides a bounded control phase and the checks that follow. GP ChatGPT ★ 488repo stars hybrid Open-RUA The hybrid prompt for Claude Code and ChatGPT guides a Go2 quadruped through a prepared indoor navigation episode. The agent chooses movement from RGB views and ends the run only by calling finish. GP ChatGPT ★ 488repo stars hybrid-search-implementation Seth Hobson The hybrid-search-implementation skill combines vector and keyword search into one retrieval flow. Use it for RAG systems and search engines when one method misses important matches. CX Codex +1 10,800installs google-cloud-solution-hybrid-search-alloydb official google The google-cloud-solution-hybrid-search-alloydb skill helps design and validate AlloyDB hybrid search, combining semantic retrieval with keyword search and structured SQL filters. CX Codex +1 3,300installs postgres-hybrid-text-search tigerdata postgres-hybrid-text-search teaches AI agents how to combine BM25 keyword matching with pgvector semantic retrieval using Reciprocal Rank Fusion in PostgreSQL. CX Codex +1 411installs Jaxonomy — hybrid dynamical systems on JAX machinavitalis Jaxonomy — hybrid dynamical systems on JAX is an MCP server for Claude Code that runs hybrid-system simulations for control-loop design and trajectory optimization. CX Codex +1 ★ 22repo stars srv-d7aoqmh5pdvs7391dcqg ciprianpater srv-d7aoqmh5pdvs7391dcqg — a remote MCP server that hands your agent the NWO Robotics API: natural-language robot commands, vision-to-action inference, task planning and a ROS2 bridge. CX Codex +1 4,056installs RTK Motion Intelligence rtkmotion RTK Motion Intelligence — an MCP server that gives your agent motion capture, rehab biomechanics, location threat and truck fleet data. Free samples first, paid calls in USDC. CX Codex +1 2,335installs

From the same set41

The whole set →
batch The batch prompt for Claude Code groups independent reads and perception queries in one model response. Physical actions stay ordered, with a pause where their result needs checking. GP ChatGPT ★ 488repo stars hybrid The hybrid Claude Code prompt coordinates a humanoid as one actor. It combines upper-body targets with walking velocity while native control handles the legs and waist. GP ChatGPT ★ 488repo stars hybrid The hybrid prompt for Claude Code reports RoboDojo task progress with evidence-backed subgoals. A reviewer checks action-model proposals before execution. GP ChatGPT ★ 488repo stars hybrid The hybrid Claude Code prompt controls an arm and dexterous hand assembly in DexJoCo. It supports pick_bucket with the published single-arm model recipe. GP ChatGPT ★ 488repo stars return_background The return_background Claude Code prompt coordinates background jobs and waits. It defines queue boundaries, result checks, and recovery after failures. GP ChatGPT ★ 488repo stars main_agent This Claude Code prompt delegates visual observation while leaving physical actions and task completion to the main agent. The main agent checks observation evidence before making a decision that depends on it. GP ChatGPT ★ 488repo stars hybrid The hybrid Claude Code prompt helps choose between actor_act and direct robot motion. It guides a bounded control phase and the checks that follow. GP ChatGPT ★ 488repo stars hybrid The hybrid prompt for Claude Code and ChatGPT guides a Go2 quadruped through a prepared indoor navigation episode. The agent chooses movement from RGB views and ends the run only by calling finish. GP ChatGPT ★ 488repo stars

Project sponsors

Companies that keep this analytics open