RPent is an open framework for building embodied agents, software systems that perceive and act in the physical world while combining reasoning, memory, execution, and adaptation. It is intended for embodied-intelligence researchers working on physical tasks and related evaluations.
Borrowing it
Nothing to install: this file belongs to RLinf/RPent. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.
curl -O https://raw.githubusercontent.com/RLinf/RPent/main/.agents/skills/add-robot/SKILL.mdgit clone --depth 1 https://github.com/RLinf/RPentWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/rlinf/rpent/add-robot)<a href="https://agentmods.dev/skills/rlinf/rpent/add-robot"><img src="https://agentmods.dev/badge/skills/rlinf/rpent/add-robot/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/rlinf/rpent/add-robot"><img src="https://agentmods.dev/badge/skills/rlinf/rpent/add-robot.svg" alt="Reviewed on agentmods" width="80" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00034 | $0.00724 |
| Opus 5.5 | $0.00014 | $0.00290 |
| Sonnet 5.5 | $0.00007 | $0.00145 |
| Haiku 4.5 | $0.00003 | $0.00072 |
Grade A, and why
add-robot scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 1mo ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 63 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Add or extend a robot integration
Start with the robot guide and primitive guide. Read the current RobotSpec, registry, and runtime helpers, then the closest existing robot package. The guide uses LIBERO as a worked example; select a sibling that fits the target's actual control, observation, and lifecycle contracts.
Establish ownership and scope
Identify the required environment, action model, tools, observations, and supported entry points. Check whether RPent or RLinf already owns each capability. Prefer shared Env/VLA components and thin adapters over duplicated simulation, action conversion, success logic, or lifecycle machinery. Explain any necessary divergence in the PR.
For an existing robot, follow only the steps affected by the change. Keep
robot-specific choices in robots/<robot>/ and optional dependencies in the
appropriate extra with lazy imports.
Wire the integration
- Expose the package factories through the existing registry. Use
RobotSpecandRunConfigfor identity, prompts, CLI parsing, and runner hooks; inspect current capability fields and consumers instead of copying a stale spec. - Reuse
BaseEnvClient/BaseEnvFacadeand, when needed,BaseVLAClient/BaseVLAFacade. Match both sides of RPC routes, observation and action formats, reset/session semantics, and server-side constraints. - Connect the toolkit's schemas, handlers, primitives, and state capture. Reuse existing artifact and memory ownership; keep large observations out of tool-result text. Check actual tool exposure across supported planners.
- Build the shared
PromptBundlefrom robot-owned content. Verify rendered variables, tool names, concise instructions, and consistency with the toolkit. Follow docs-check for model-facing prose. - Use runtime helpers for owned daemons and borrowed endpoints. Cover startup failure and cleanup; handle complete CLI startup and selected Dashboard components when supported. Keep robot-specific result schemas in the robot's finalization hook and verify which entry points call it.
- Update the relevant installation extra, usage/development documentation, and tests. Follow the actual runner and CI configuration when adding a new supported test target; adding a directory alone does not wire it into CI.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 1mo ago First seen · 63 lines · 34 tokens per session scan A 449a9c90cf1a
add-robot is a skill published in the GitHub repository RLinf/RPent (1,640 stars, last pushed 3d ago), licensed Apache-2.0. It adds 34 tokens to every session and 724 once invoked, about $0.0001 per session on Opus 5.5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-09-17.
Other skills, from other repositories
spark-environment-setup
Set up a working ML training/inference environment on NVIDIA DGX Spark (GB10, aarch64, CUDA 13). Use when installing PyTorch/Unsloth/TRL/vLLM on DGX Spark, hitting libcudart or wheel-ABI errors on aarch64, or choosing between NGC containers and bare pip installs.
spark-memory-thermal-ops
Manage unified memory and thermals during long-running ML jobs on NVIDIA DGX Spark. Use when planning memory headroom for a training run on GB10, when a job OOMs on unified memory, or when monitoring temperature and power during multi-hour training.
spark-training-gotchas
Preflight and diagnose the ten known failure modes for ML training on NVIDIA DGX Spark. Use when a training run on DGX Spark fails to start, OOMs below the 128GB limit, slows down mid-run, or before any multi-hour training job on GB10.
terraform-module-library
Build reusable Terraform modules for AWS, Azure, GCP, and OCI infrastructure following infrastructure-as-code best practices. Use when creating infrastructure modules, standardizing cloud provisioning, or implementing reusable IaC components.
cdf-simulation
Component-owned guidance for CDF Simulation requests submitted directly or through Airflow-triggered DIG Flow Catalog prompts — golden THT board, THT pin-geometry variation, good/clean SMT images, structural pose defects, and solder texture defects, from a board USD. Turns one matched prompt into one validated…
amc-run-rtsp-calibration
Calibrate a new dataset from live RTSP camera streams via the AutoMagicCalib REST API. Use when the user provides RTSP URLs or asks to calibrate live cameras; VIOS records clips, AMC ingests them, then runs calibration.