Skip to main content

Kite CLI and MCP server

Robot learning on cloud GPUs, from your terminal or your AI agent. Kite trains robot policies with reinforcement learning in simulation, evaluates policies on Kite SDK specs, augments LeRobot datasets, and rebuilds robot episodes as MuJoCo digital twins. This package gives you the kite command line and the kite-mcp server for Claude, Cursor, and any Model Context Protocol (MCP) client.

  • RL training in simulation. Describe a behavior in plain English. Kite writes the task spec, checks its rewards for free, and trains the policy on a GPU. You get the policy as ONNX, the MuJoCo scene it trained in, and clips.
  • RL verdicts. Every RL run ends with a report: a pass, needs_review or fail verdict from measured checks (falls per minute, survival, command tracking, gait, posture), plus a vision model's read of the clip.
  • Dataset augmentation. Relight a LeRobot dataset on Hugging Face or in a public S3 bucket to match a deployment scene, or generate new variations of it with video augmentation. Output is a standard LeRobot dataset (Parquet + MP4), downloaded or pushed to Hugging Face.
  • Digital twins (beta). Point Kite at one episode of a LeRobot dataset and get an interactable MuJoCo scene of the room, with the objects the robot handles built to size.
  • Evaluations (beta). Evaluate a policy by running a Kite SDK spec on Kite's machines, on an NVIDIA L4 for a GPU policy: every episode traced, every evaluator scored. You get a scorecard and the run directory, which kite verify checks.

Install

pip install "kiteml-cli[mcp]"

Python 3.10 or later. Drop [mcp] if you only want the CLI.

Sign in

kite auth login     # opens the dashboard; stores a 90-day API key
kite doctor         # checks the API is reachable and your key works

On a server or in CI, set an API key from app.kiteml.com instead:

export KITE_API_KEY=kite_...

Use it from an AI agent

Kite hosts the MCP server, so there's nothing to run:

claude mcp add --transport http kite https://mcp.kiteml.com/mcp

Or run it locally over stdio with this package:

claude mcp add kite -e KITE_API_KEY=$KITE_API_KEY -- kite-mcp

For Cursor and other clients, add {"mcpServers": {"kite": {"url": "https://mcp.kiteml.com/mcp"}}} to the MCP config. Then ask, for example: "train the Open Duck Mini to walk forward and tell me when it passes", or "relight lerobot/pusht to look like this photo".

Tools What they do
kite_rl_catalog, kite_rl_plan, kite_rl_validate See which robots and objectives RL runs support, and plan and check a training spec for free
kite_rl_train, kite_rl_fork, kite_rl_cancel Start, branch from, and stop RL runs
kite_rl_status, kite_rl_metrics, kite_rl_report, kite_rl_list Follow training and read each run's verdict
kite_augment_create, kite_augment_status, kite_augment_list, kite_augment_cancel Augment LeRobot datasets
kite_twin_validate, kite_twin_create, kite_twin_status, kite_twin_list, kite_twin_cancel, kite_twin_resume Build MuJoCo digital twins (beta)
kite_eval_validate, kite_eval_run, kite_eval_status, kite_eval_list, kite_eval_report, kite_eval_logs, kite_eval_cancel, kite_eval_resume Evaluate a policy on a Kite SDK spec on Kite's machines and read the scorecard (beta)
kite_doctor Check connectivity and authentication

Use it from the terminal

Train a walking policy for the Open Duck Mini v2:

kite rl plan open_duck_mini_v2 "walk forward at a steady pace" -o walk.json
kite rl validate walk.json                 # free: what each reward term pays canned policies
kite rl train walk.json --budget probe --wait
kite rl report rlr_...                     # the verdict and each check behind it
kite rl download rlr_...                   # -> kiteml_rlr_.../policy/policy.onnx

Augment a dataset:

kite augment create --repo-id lerobot/pusht \
  -i "change the table surface to white marble, vary the lighting" -n 20 --wait
kite augment download aug_... -o ./pusht-marble

--repo-id takes a Hugging Face id (lerobot/pusht), its huggingface.co/datasets/ URL, or an S3 link to the dataset's folder in a public bucket (s3://bucket/dataset/).

Build a digital twin (beta, enabled per account):

kite twin create lerobot/svla_so101_pickplace --robot so101 --out ./twins   # -> ./twins/twin_.../scene.xml

Evaluate a policy on a Kite SDK spec, on Kite's machines (beta):

kite eval validate specs/smoke.yaml            # free: the plan and what a run reserves
kite eval run specs/smoke.yaml --out ./runs    # -> ./runs/eval_.../scorecard.json

A file too big to send, such as a checkpoint's weights, can come from the Hugging Face Hub or a public S3 bucket: --hf PATH=OWNER/NAME@COMMIT:FILE or --hf PATH=<the file's huggingface.co URL>, and --s3 PATH=s3://BUCKET/KEY, PATH being the path the spec names. A checkpoint of one of your Kite training runs goes as --training-run PATH=trn_...[@STEP] (its latest without @STEP), PATH being the checkpoint folder the spec names: Kite reads it from its own storage and publishes nothing. One of your twins goes as --twin PATH=twin_..., PATH being the folder the twin unpacks to; the files in your PATH that differ from the twin's go inline beside it.

Every command prints JSON ({"ok": true, "data": ...}) so scripts and agents can parse it. Run kite --help for the full list.

Metadata

Release files for kiteml-cli 0.8.0

For a detailed explanation of source distributions (sdists) and built distributions (wheels), please see the package formats documentation.

Source distribution (sdist)

Source distribution for kiteml-cli 0.8.0
File Size Uploaded Provenance Actions
kiteml_cli-0.8.0.tar.gz 46.0 kB With provenance Details for kiteml_cli-0.8.0.tar.gz

Built distribution (wheel)

Table of built distributions (wheels) for kiteml-cli 0.8.0
File Interpreter ABI Platform Actions
kiteml_cli-0.8.0-py3-none-any.whl Python 3 none any Details for kiteml_cli-0.8.0-py3-none-any.whl

Total release size: 102.2 kB

Release files / kiteml_cli-0.8.0.tar.gz

Download URL kiteml_cli-0.8.0.tar.gz
Size 46.0 kB
Tags Source
SHA-256 checksum
How to use checksums
9643705a607596de3fc5fd1019bfad6444b4b192869e0d621c33a6d0172c4bbb
BLAKE2b-256 checksum
How to use checksums
f8b39a358a7ccadb9414fbf27ecf8d64e3bbe557204c8123177382b9f99a0cbb
Upload date
Uploaded using Trusted Publishing?
What is trusted publishing?
Yes
Uploaded via twine/7.0.0 CPython/3.13.14

Provenance

Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.

PyPI Publish Attestation

PyPI verified that this artifact, at this checksum, originated from the publisher listed below.

Signed by GitHub Actions, verified by PyPI on Oct 27, 2026.

Transparency log

Release files / kiteml_cli-0.8.0-py3-none-any.whl

Download URL kiteml_cli-0.8.0-py3-none-any.whl
Size 56.2 kB
Tags Python 3
SHA-256 checksum
How to use checksums
2b25555122a27098b7846feea36ce4d1890e4f18fbac3343d4805c61247e1294
BLAKE2b-256 checksum
How to use checksums
69b3f835dcf381b1020eeb69a80bd03a73c4d1bce5c8ea5118f3371e009bc3b2
Upload date
Uploaded using Trusted Publishing?
What is trusted publishing?
Yes
Uploaded via twine/7.0.0 CPython/3.13.14

Provenance

Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.

PyPI Publish Attestation

PyPI verified that this artifact, at this checksum, originated from the publisher listed below.

Signed by GitHub Actions, verified by PyPI on Oct 27, 2026.

Transparency log

Security for kiteml-cli 0.8.0

Provenance

Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.

Full provenance

Both files in this release have been uploaded with provenance metadata.

Select from 2 files

Attestation for kiteml_cli-0.8.0.tar.gz

PyPI Publish Attestation

PyPI verified that this artifact, at this checksum, originated from the publisher listed below.

Signed by GitHub Actions, verified by PyPI on Oct 27, 2026.

Transparency log

A note on package security

PyPI is a repository for distributing packages, not a service for auditing them. While third-party security researchers and our staff do their best to detect and remove malicious packages, PyPI cannot guarantee the quality, security or suitability of any given package.

To learn more about protecting yourself from malicious actors, visit our package security documentation.

Anthropic, PBC Visionary sponsor Bloomberg Visionary sponsor Hudson River Trading Visionary sponsor Meta Visionary sponsor NVIDIA Visionary sponsor Microsoft Sustainability sponsor Depot Continuous Integration AWS Cloud computing and Security Sponsor Datadog Monitoring Fastly CDN Google Download Analytics Sentry Error logging StatusPage Status page