Skip to main content

labagent: CLI to launch hands-on labs for RAG, text-to-SQL, Databricks, AWS and Azure

PyPI Python CI License

labagent builds a lab with one command and removes it with one more. Each lab is a pack: ordered, re-runnable steps. launch resumes after a failure. destroy removes only what the pack created.

PIP setup On Mac

brew install pipx && pipx ensurepath
pipx install labagent

PIP setup on Windows

# 1. Install prerequisites & labagent
winget install Python.Python.3.14
Invoke-WebRequest -Uri https://bootstrap.pypa.io/get-pip.py -OutFile get-pip.py
python get-pip.py
pip install labagent                             # Mac: brew install pipx && pipx ensurepath && pipx install labagent
labagent packs                                   # list labs (add --json for a machine-readable catalog)
labagent launch <pack> --name demo --param k=v   # build or resume
labagent destroy <pack> --name demo              # remove what launch created

Extras: labagent[rag] (Databricks RAG), [rag-local] (laptop RAG), [rag-lab] (laptop RAG with JupyterLab), [aml], [modal]. Python 3.9+. On a Mac, use pipx (above): plain pip install puts labagent in ~/Library/Python/3.X/bin, which is not on your PATH. Add --dry-run to any launch to see the steps.

Packs

Pack What it builds Where
azure-labs Copy Azure VMs, disks and network from one resource group to another with snapshots, or snapshot a dead VM or disk. Source is read-only. An Azure Lab Services alternative. Guide Azure (needs az login)
rag-sec-tutorial Nine notebooks: RAG on SEC 10-K filings with embeddings, pgvector, hybrid search, citations, an evaluation gate, a deployed app Databricks Free Edition
db-agent db-agent text-to-SQL app on Databricks Apps, pinned commit. Which model, warehouse and schema it picks Databricks Free Edition
dbx-lakehouse Medallion lakehouse (bronze, silver, gold) from real open data, with masks, a row filter, notebooks and exercises. Guide Databricks Free Edition, AWS, Azure
dbx-workspace Premium Databricks workspace via Terraform. Plan only unless apply=yes. Guide AWS or Azure (costs money)
aml Synthetic anti-money-laundering lab with an answer key: RAG for developers, Microsoft 365 Copilot for staff. Guide Your machine
rag-ide Private browser VS Code per learner with the RAG lab Modal, EC2 or Docker

Also: labagent rag init --dir ~/rag-lab makes a laptop RAG lab of nine notebooks (needs [rag-lab]), and labagent rag destroy --dir ~/rag-lab removes it. See docs/rag-local.md. AWS cohort sandboxes (Terraform IAM, plan-then-apply cleanup) run through labagent run --job job.json: see docs/job-contract.md.

Azure: move VMs between resource groups

export LABAGENT_ALLOWED_AZURE_SUBSCRIPTIONS=$(az account show --query id -o tsv)
labagent launch azure-labs --name move1 --param source_rg=legacy-rg --param target_rg=new-rg
labagent launch azure-labs --name old1 --param source_rg=legacy-rg --param target_rg=archive-rg --param snapshot_only=true
labagent destroy azure-labs --name move1 --param target_rg=new-rg

Parameters: source_rg, target_rg, vms (comma list, default all), snapshot_only, subscription. Same subscription and region. It does not copy availability sets, scale sets, unmanaged VHDs, customer-key encryption, extensions or public IP addresses (a new address is created). Tested on a real subscription with Ubuntu VMs. Not Azure Lab Services and not from Microsoft.

For AI coding agents and scripts

  • Output is one JSON envelope on stdout (status: ok, partial, failed; items per step). Exit code 0, 1, 2. Logs go to stderr and never contain secrets.
  • labagent packs --json lists every pack with parameters, defaults, platforms and cost tier.
  • Every launch is idempotent. Run it again to resume. destroy deletes only resources carrying the pack's tags or state.
  • Safety gates: LABAGENT_ALLOWED_ACCOUNTS (AWS), LABAGENT_ALLOWED_SUBSCRIPTIONS (Databricks on Azure), LABAGENT_ALLOWED_AZURE_SUBSCRIPTIONS (azure-labs). An empty list refuses to run.
  • Plain English front end: labagent "deploy rag-tutorial" prints the command it understood and asks first (local Ollama or any OpenAI-compatible endpoint).

Status

Run end to end on real services: Databricks Free Edition labs, the AWS premium workspace, and azure-labs copy and delete on Azure. Tested only against mocks so far: Azure and custom-network dbx-workspace, and dbx-lakehouse on AWS and Azure. Try labs in a throwaway workspace or account first.

Apache-2.0. SECURITY.md, CONTRIBUTING.md, docs/packs.md. Related: db-agent, labagent-datasets. Maintained by BeCloudReady.

Metadata

Release files for labagent 0.4.0

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

Source distribution (sdist)

Source distribution for labagent 0.4.0
File Size Uploaded Provenance Actions
labagent-0.4.0.tar.gz 424.8 kB With provenance Details for labagent-0.4.0.tar.gz

Built distribution (wheel)

Table of built distributions (wheels) for labagent 0.4.0
File Interpreter ABI Platform Actions
labagent-0.4.0-py3-none-any.whl Python 3 none any Details for labagent-0.4.0-py3-none-any.whl

Total release size: 784.2 kB

Release files / labagent-0.4.0.tar.gz

Download URL labagent-0.4.0.tar.gz
Size 424.8 kB
Tags Source
SHA-256 checksum
How to use checksums
49d20347efa1ac0660bd3b4a05e3208e50d113bc02050360b46ce71e612ac5aa
BLAKE2b-256 checksum
How to use checksums
acf357eee8fda8e481991c17714f88c3cacb254b050e229e7c28139eb0e3ff83
Upload date
Uploaded using Trusted Publishing?
What is trusted publishing?
Yes
Uploaded via twine/7.0.0 CPython/3.13.16

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 Nov 4, 2026.

Transparency log

Release files / labagent-0.4.0-py3-none-any.whl

Download URL labagent-0.4.0-py3-none-any.whl
Size 359.5 kB
Tags Python 3
SHA-256 checksum
How to use checksums
eda51b1e3cfd3423d488792e8c1884edf5c572e0a6be774362ef5b1366432cd6
BLAKE2b-256 checksum
How to use checksums
68a4c8384b121611bffec85e5d0bfd0d4d01944d6d90ca82a8c1e835da06c8ed
Upload date
Uploaded using Trusted Publishing?
What is trusted publishing?
Yes
Uploaded via twine/7.0.0 CPython/3.13.16

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 Nov 4, 2026.

Transparency log

Security for labagent 0.4.0

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 labagent-0.4.0.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 Nov 4, 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