# tensorchord/envd

🏕️ Reproducible development environment for humans and agents

Repository: https://github.com/tensorchord/envd
Canonical: https://ross.abutalabs.com/products/envd
Homepage: https://envd.tensorchord.ai/
Language: Go
License: Apache-2.0
License Family: permissive
Topics: developer-tools, development-environment, docker, buildkit, hacktoberfest, llmops, mlops, mlops-workflow, model-serving, agent, code-agent, codex
Last push: 2026-07-25T03:58:33+00:00

## Health v2 (maintenance only)
Score: 86/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 94, release rhythm 69, longevity 100
- inputs: {"age_days": 1605, "days_push": 39, "days_rel": 208, "gap_med": 1.5, "n_releases_24m": 21}
- flags: none
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 2227, forks 168 (observed 2026-08-28T04:06:28.833681+00:00)

## What it is
envd is a command-line tool that creates reproducible, container-based development environments for AI/ML work using a simple Python-like build.envd declaration instead of hand-written Dockerfiles. It builds OCI-compatible images with built-in support for Jupyter, VSCode remote, CUDA, and conda/pip dependencies, and can also provision environments for AI coding agents.

## Use cases
- set up a reproducible python machine learning dev environment
- replace dockerfiles with a declarative build.envd
- create a container with cuda and pytorch dependencies
- spin up jupyter and vscode remote in an isolated container
- share and publish dev environments as OCI images
- provision consistent environments for AI coding agents like codex
- reuse environment build functions across a team via git includes

## When to choose
- you do AI/ML development with complex Python, CUDA, and system dependencies
- you want reproducible, versioned dev environments without writing Dockerfiles
- your team needs to share and reuse environment definitions
- you want first-class Jupyter and VSCode remote support in containers

## When to avoid
- you need a general-purpose CI/CD pipeline tool rather than a dev environment manager
- your project has no containerization or dependency complexity
- you cannot use Docker or an OCI-compatible runtime
- you need lightweight virtualenv/conda-only workflows without containers

## Facets
- artifact type: cli-tool
- maturity: active
- function: developer-tools, container-runtime, build-tool, configuration-management, deployment
- domain: developer-tools, machine-learning, data-science
- platform: windows, cli, python, go
- tags: development-environment, buildkit, reproducible-environments, mlops, llmops, cuda, jupyter, vscode-remote, oci-images, ai-agents, containers, command-line, linux, macos, docker

## Member repositories
- tensorchord/envd (main) score 86

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:06:28.833681+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-30T02:44:51.746900+00:00, confidence not recorded.
  - readme: https://github.com/tensorchord/envd (fetched 2026-08-28T04:06:28.833681+00:00, sha 5755faf29fff)
  - homepage: https://envd.tensorchord.ai/ (fetched 2026-08-29T10:25:19.576403+00:00, sha 6fd55d97a612)
  - site_page: https://envd.tensorchord.ai/guide/getting-started (fetched 2026-08-29T10:25:19.586320+00:00, sha 3e2c382b7f83)
- Data as of 2026-08-30T08:39:29.467469+00:00.