# superagent-ai/vibekit

Run Claude Code, Gemini, Codex — or any coding agent — in a clean, isolated sandbox with sensitive data redaction and observability baked in.

Repository: https://github.com/superagent-ai/vibekit
Canonical: https://ross.abutalabs.com/products/vibekit
Homepage: https://vibekit.sh
Language: TypeScript
License: MIT
License Family: permissive
Topics: agent, ai, claude-code, codex, vibe-coding, gemini-cli
Last push: 2026-01-13T18:48:57+00:00

## Health v2 (maintenance only)
Score: 49/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 62, release rhythm 41, longevity 34
- inputs: {"age_days": 476, "days_push": 232, "days_rel": 394, "gap_med": 1, "n_releases_24m": 38}
- flags: none
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 1851, forks 246 (observed 2026-08-28T04:05:44.363513+00:00)

## What it is
VibeKit is a CLI tool that wraps AI coding agents such as Claude Code, Gemini CLI, Grok CLI, and Codex in isolated Docker sandboxes with automatic sensitive-data redaction. It also ships an SDK and auth package so applications can execute AI-generated code safely with built-in observability.

## Use cases
- run claude code in an isolated sandbox
- execute AI-generated code safely on my machine
- redact secrets and API keys from AI agent output
- add observability and tracing to coding agent sessions
- prevent coding agents from touching my local environment
- run gemini cli or codex offline without cloud dependencies
- embed secure coding agent execution into my app

## When to choose
- you use coding agents like Claude Code, Gemini CLI, or Codex and want them sandboxed away from your local setup
- you need automatic redaction of secrets and API keys from agent completions
- you want real-time logs, traces, and metrics for what agents are doing
- you prefer a fully local, offline workflow with no cloud dependency
- you are building an application that must execute AI-written code securely via an SDK

## When to avoid
- you need a framework to build your own agents from scratch rather than a safety wrapper around existing ones
- you require centralized, cloud-hosted or team-wide sandbox enforcement instead of local Docker isolation
- your environment cannot run Docker or Node.js/npm
- you only need a coding agent itself, not sandboxing or redaction around it

## Facets
- artifact type: cli-tool
- maturity: active
- function: security, monitoring, tracing, logging, cli
- domain: developer-tools, security, large-language-models
- platform: cli, cross-platform, windows
- tags: sandbox, coding-agents, claude-code, gemini-cli, codex, secrets-redaction, data-redaction, observability, vibe-coding, ai-safety, docker-sandbox, agent-wrapper, sdk, sandboxing, ai-agents, containers, command-line, nodejs, docker, linux, macos

## Member repositories
- superagent-ai/vibekit (main) score 49

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:05:44.363513+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 2026-08-30T03:16:56.116527+00:00, confidence not recorded.
  - readme: https://github.com/superagent-ai/vibekit (fetched 2026-08-28T04:05:44.363513+00:00, sha 6a8b5b9ea0b4)
- Data as of 2026-08-30T08:39:29.467469+00:00.