Agentic Robot Creation through Generation, Verification, and Revision

Xuebing Li; Yibo He; Sitong Liu; Ran Ji; Jiayuan Gu

cs.ROOct 13, 2026

Agentic Robot Creation through Generation, Verification, and Revision

Authors: Xuebing Li, Yibo He, Sitong Liu, Ran Ji, Jiayuan Gu

Organizations: ShanghaiTech University

Abstract

Can an AI agent turn high-level user intent into a functioning robot and teach it how to behave? We study this question from two complementary aspects: embodiment and behavior creation. For embodiment creation, we investigate how well current multimodal models capture mechanical requirements, and introduce computational metrics for evaluating generated designs before fabrication. For behavior creation, we present Roboforge, an agentic framework that turns task specifications into executable policies through motion generation, policy learning, and execution-grounded revision. Given a robot embodiment and task, the agent generates a reference motion, configures the learning objective, and trains a policy. Crucially, motion remains in the loop: rollout feedback can revise the reference motion, reward, or both. We evaluate a generated embodiment across multiple MLLMs and behavior creation across four diverse robots against reward-refinement and one-shot motion baselines. Our computational metrics expose recurring mechanical failures in generated designs, while behavior experiments show that execution-grounded refinement improves task-level behavior and execution quality. Finally, we fabricate an agent-generated Gorilla and directly deploy learned behaviors on the physical robot.

Figures & tables

Check category	GPT6-Astra	GPT5.6-Sol	GPT5.6-Terra	GPT 5.5
Design and Assembly	6/10	4/10	4/10	3/10
Mechanical and Actuation Feasibility	4/6	0/6	1/6	0/6
Motion and Task Execution	5/7	1/7	1/7	1/7

Table 1: Pre-fabrication feasibility of robot embodiments generated by GPT variants. Each entry

x/y

denotes the number of passed checks among the

y

checks in that category. Appendix I.2 provides the complete check definitions and per-design verdicts.

Figure 1: Generated robots (top) and zoomed details (bottom). GPT6-Astra shows the correct mounting-hole layout. The other views expose intersections between the forearm and rear leg in GPT5.6 Sol, the torso and upper arm in GPT 5.5, and the sideplate and elbow motor in GPT5.6 Terra. Surrounding parts are hidden or made transparent to reveal the interfaces.

Figure 2: Overview of RoboForge: motion-guided, execution-grounded policy learning. Given a robot embodiment and task, the agent generates and validates a reference motion, configures the learning objective, and trains a policy. Execution feedback then refines the motion, reward, or both while keeping the embodiment and task fixed.

Method	Go2	MicroDuck	Ostrich	Gorilla	Overall
Eureka-style	2.79	2.01	2.00	1.59	2.10
One-shot Motion	3.93	3.30	4.04	3.14	3.60
RoboForge (Ours)	4.10	4.03	3.99	3.77	3.97

Table 2: User-study ratings across four robot embodiments (

n=57

;

\uparrow

). RoboForge achieves the highest overall rating and the highest rating on three of four embodiments.

Method	Yaw Err. (rad)	Foot-slip (m/s)	Action Var. (rad 2 )	Joint Acc. (rad/s 2 )
Eureka-style	0.0345	0.121	0.0830	114
One-shot Motion	0.0584	0.0716	0.00665	31.7
RoboForge (Ours)	0.00836	0.0131	0.00290	22.0

Table 3: Reference-independent execution-quality metrics (

\downarrow

). RoboForge achieves the lowest mean value for every metric.

Figure 3: Qualitative comparison of learned policy executions on the gorilla handstand task. Eureka-style fails after an initial rear-leg lift, while One-shot loses balance with outward leg rotation. RoboForge folds the legs inward before inversion, enabling a balanced transition and successful completion.

	Agent Behavior		Final Policy Quality
Method	Motion Gen.	Motion Refine	User Rating $\uparrow$	Yaw Err. $\downarrow$	Foot-slip $\downarrow$	Action Var. $\downarrow$	Joint Acc. $\downarrow$
GPT	16/19	7/19	3.61	0.0172	0.0419	0.0179	29.2
RoboForge (Ours)	19/19 ∗	11/19	3.97	0.00836	0.0131	0.00290	22.0

Table 4: Comparison with unguided GPT. GPT spontaneously introduces and refines reference motions, while our structured pipeline achieves higher user ratings and better execution metrics. ∗ Motion generation is prescribed in our pipeline.

Figure 4: Motion refinement preserves task semantics while improving gait realization. The initial gait has two airborne feet (a), whereas the refined gait uses single-foot swing (b) with coordinated body shift, yielding a more stable four-beat gait.

Figure 5: Refined motions from GPT and RoboForge. Both increase swing-foot clearance, but differ in whole-body coordination. CoM trajectories show that GPT remains nearly straight, whereas RoboForge shifts laterally toward the supporting legs for coordinated weight transfer.

Figure 6: Physical realization of the generated Gorilla and its learned behaviors. Top: hand waving. Bottom: walking. Frames are ordered from left to right.

Appendix figures & tables14 assets

Supplementary material from the paper’s appendix.

Appendix

Panel of Figure 7 — Figure 7: GPT-generated robot embodiments used in our experiments. (a) Ostrich. (b) Gorilla, whose design is additionally validated through physical fabrication and assembly.

Figure 8: Fabricated Gorilla embodiment.

Robot	Physics step (s)	Decimation	Policy (Hz)
Go2, MicroDuck, Ostrich	0.005	4	50
Gorilla	0.002	10	50

Appendix

Table 5: Simulation timing shared across methods. All robots are controlled at a policy frequency of 50 Hz.

Parameter	Default value
Parallel environments	$4{,}096$
Policy steps per environment per iteration	$24$
Transitions per iteration	$98{,}304$
Optimization epochs per iteration	$5$
Minibatches per epoch	$4$
Transitions per minibatch	$24{,}576$

Appendix

Table 6: Default PPO and rollout configuration of the shared RL template.

Robot	Walking tasks	Whole-body skills	Total
Go2	High-step walk; low walk	Bow; look around; kick	5
MicroDuck	Fast walk; waddling walk	Bow; look around; kick	5
Ostrich	Happy walk; sad walk	Nod; stretch; squat and rise	5
Gorilla	Alternating walk	Handstand; stretch; wave	4
Total	7 walking tasks	12 skills	19

Appendix

Table 7: Task inventory across four embodiments: seven walking tasks and twelve whole-body skills.

Figure 9: Generated stylized locomotion motions for Ostrich. Left: happy walking ; right: sad walking . Both motions are generated by our method and share the same forward-locomotion objective, while differing in head–neck posture and its coordination with the stepping pattern.

Figure 10: Generated motion for MicroDuck’s kick task produced by our method. The sequence coordinates gaze, lateral body shift, the kicking leg, and the supporting leg across preparation, single-foot support, and recovery.

Score	Description
1	The task is largely unfulfilled, or the motion has severe problems.
2	Few task requirements are fulfilled, or motion problems are substantial.
3	The task is partially or mostly fulfilled, with noticeable shortcomings.
4	The task is fulfilled with generally stable motion and minor problems.
5	The task is fulfilled with smooth, coordinated, and stable motion.

Appendix

Table 8: Five-point rating rubric used in the user study.

Robot	Method	$n_{W}$	Walking	$n_{S}$	Non-walking	Overall
Go2	Eureka-style	29	$3.00\pm 1.54$	43	$2.65\pm 1.38$	$2.79\pm 1.44$
	One-shot	29	$3.00\pm 1.44$	43	$\mathbf{4.56}\pm 0.63$	$3.93\pm 1.28$
	RoboForge (Ours)	29	$\mathbf{3.76}\pm 0.91$	43	$4.33\pm 0.78$	$\mathbf{4.10}\pm 0.87$
MicroDuck	Eureka-style	28	$2.64\pm 1.34$	43	$1.60\pm 0.85$	$2.01\pm 1.18$
	One-shot	28	$2.82\pm 1.09$	43	$3.60\pm 1.29$	$3.30\pm 1.27$
	RoboForge (Ours)	28	$\mathbf{4.00}\pm 1.02$	43	$\mathbf{4.05}\pm 0.92$	$\mathbf{4.03}\pm 0.96$

Appendix

Table 9: User ratings by robot and task group (

\uparrow

), reported as mean

\pm

sample standard deviation.

n_{W}

and

n_{S}

denote the numbers of valid individual ratings per method for walking and non-walking tasks, respectively, and are not counts of distinct participants. Bold indicates the highest unrounded mean among the displayed methods within each robot and column, including ties.

Robot	Method	Yaw Err.	Action Var.	Joint Acc.	Foot-slip
		(rad)	(rad 2 )	(rad/s 2 )	(m/s)
Go2	Eureka-style	$\mathbf{0.00807}$	$0.0531$	$72.7$	$0.148$
	One-shot	$0.063$	$0.00307$	$23.3$	$0.2$
	RoboForge (Ours)	$0.0135$	$\mathbf{0.000174}$	$\mathbf{13.9}$	$\mathbf{0.0354}$
MicroDuck	Eureka-style	$0.0779$	$0.034$	$122$	$0.0564$
	One-shot	$0.139$	$0.0222$	$77.9$	$0.0674$

Appendix

Table 10: Physical execution metrics by robot (

\downarrow

). Yaw error and slip are averaged over walking tasks, whereas action variation and joint acceleration are averaged over all tasks. Values are rounded to three significant digits. Bold indicates the minimum among the displayed methods within each robot and metric, including ties. Action var. denotes second-order action variation normalized by the number of named joints.

Robot	Method	Walking	Non-walking
Go2	GPT	$3.14\pm 1.36$	$3.86\pm 1.15$
	RoboForge (Ours)	$\mathbf{3.76}\pm 0.91$	$\mathbf{4.33}\pm 0.78$
MicroDuck	GPT	$3.61\pm 1.13$	$3.91\pm 1.06$
	RoboForge (Ours)	$\mathbf{4.00}\pm 1.02$	$\mathbf{4.05}\pm 0.92$
Ostrich	GPT	$3.21\pm 1.10$	$4.05\pm 0.83$
	RoboForge (Ours)	$\mathbf{3.57}\pm 0.96$	$\mathbf{4.25}\pm 0.84$

Appendix

Table 11: User ratings on walking and non-walking tasks (

\uparrow

), reported as mean

\pm

sample standard deviation. Bold indicates the higher mean between GPT and RoboForge within each robot and task group.

Robot	Method	Yaw Err.	Action Var.	Joint Acc.	Foot-slip
		(rad)	(rad 2 )	(rad/s 2 )	(m/s)
Go2	GPT	$\mathbf{0.00746}$	$0.0698$	$68.9$	$0.0905$
	RoboForge (Ours)	$0.0135$	$\mathbf{0.000174}$	$\mathbf{13.9}$	$\mathbf{0.0354}$
MicroDuck	GPT	$0.0133$	$\mathbf{0.00106}$	$\mathbf{22.3}$	$0.0241$
	RoboForge (Ours)	$\mathbf{0.00435}$	$0.0102$	$46.4$	$\mathbf{0.0104}$
Ostrich	GPT	$0.0143$	$\mathbf{0.00054}$	$17.7$	$0.0355$

Appendix

Table 12: Physical execution metrics of GPT and RoboForge across robot embodiments (

\downarrow

). Yaw error and slip are averaged over walking tasks, whereas action variation and joint acceleration are averaged over all tasks. Bold indicates the lower value.

Figure 11: Final learned policy executions for the Go2 low-walk task. Figure 5 shows the corresponding reference-motion revisions, whereas this figure shows the final policies learned from those refined motions. RoboForge more clearly realizes the requested crouched gait, with distinct stepping, sustained low posture, and coordinated support transfer across the gait cycle.

ID	Check	GPT6 Astra	GPT5.6 Sol	GPT5.6 Terra	GPT 5.5
Design and assembly
DA01	Solid validity	Pass	Pass	Pass	Fail
DA02	Manufacturing mesh	Pass	Fail	Fail	Fail
DA03	Export volume fidelity	Pass	Pass	Pass	Pass
DA04	BOM completeness	Pass	Pass	Pass	Pass
DA05	Interface geometry	Pass	Fail	Fail	Fail

Appendix

Table 13: Pass/Fail results for all robot-generation checks. The category counts in the main paper can be recomputed directly from this table.

Agentic AI for Robotics

See all 39 papers

RAPID: Robot Agentic Programming from Demonstrations
Sep 24, 2026Yuyao Liu, Jiayuan Mao, David Hsu +2Program Synthesis Contact-Rich Robotic Manipulation
Massachusetts Institute of Technology · National University of Singapore · University of Pennsylvania +1
iAm.md: Robot Skill Self-Assessment through Agentic Introspection for Unknown Open-Vocabulary Domains
Oct 7, 2026Vincenzo Guarino, Emanuele Musumeci, Vincenzo Suriani +1Open-Vocabulary Semantic Mapping Agentic AI for Robotics
Department of Computer, Control and Management Engineering “Antonio Ruberti”, Sapienza University of Rome, Via Ariosto 25, 00185 Rome, Italy
Agent as Policy for Robotic Manipulation
Sep 11, 2026Mengzhao Jia, Yang Lin, Xixin Zhang +3Robotic Control Robot Policy Learning
University of Notre Dame · University of California San Diego · San Diego State University

Sep 24, 2026J/K move · Enter open · S save

RAPID: Robot Agentic Programming from Demonstrations

Yuyao Liu, Jiayuan Mao, David Hsu, Leslie Pack Kaelbling+1

Massachusetts Institute of Technology · National University of Singapore · University of Pennsylvania+1

Program Synthesis Contact-Rich Robotic Manipulation Robot Learning Generalization in Robotic Manipulation Code Generation for Robotics Agentic AI for Robotics+1

Open paper AskPDF arXiv

Coding agents have demonstrated enormous success in solving complex programming problems. To leverage their potential for robot systems, this work introduces Robot Agentic Programming from Demonstrations (RAPID), which automatically generates, verifies, and refines robot programs, given a single visual human demonstration. The iterative agentic loop of code refinement requires several key ingredients: (i) a testable task specification, (ii) action primitives for robot execution, and (iii) an interactive environment for program execution and verification. RAPID infers all three from the demonstration automatically. To make the resulting program reusable beyond the demonstration setting, RAPID uses an object-centric relational program representation that focuses on the underlying structure of the demonstrated strategy rather than the specific motion per se: it expresses the action primitives as trajectory-optimization programs that realize object-level motion effects, while composing them through relational constraints that capture scene-specific geometry at run time. We evaluated RAPID in simulation on eight challenging contact-rich nonprehensile manipulation tasks as well as general prehensile manipulation tasks in the LIBERO-Pro benchmark. We also successfully deployed it on a real Franka arm and evaluated on all eight nonprehensile tasks. In all experiments, RAPID demonstrated strong performance, with generalization over object pose, shape, material, and environment. Website: https://yuyaoliu.me/projects/rapid.

Abstract

Agentic AI for Robotics

Explore similar work

RAPID: Robot Agentic Programming from Demonstrations

iAm.md: Robot Skill Self-Assessment through Agentic Introspection for Unknown Open-Vocabulary Domains

Agent as Policy for Robotic Manipulation