generated: '2026-08-06'
method: searched
source: https://registry.npmjs.org/ + https://pypi.org/ + https://docs.activefence.com/
notes: >-
  ActiveFence now publishes under the Alice brand. First-party client libraries are published to
  npm under the @alice-io scope and to PyPI under bare names, and are linked from the SDK section
  of docs.activefence.com. There is no Maven Central, NuGet, Go module, RubyGems, Packagist or
  crates.io package for this provider — those registries were probed and returned nothing
  first-party.
packages:
- language: python
  registry: pypi
  name: wonderfence-sdk
  version: 2.1.1
  url: https://pypi.org/project/wonderfence-sdk/
  install: pip install wonderfence-sdk
  repository: https://github.com/ActiveFence/activefence_client_sdk
  requires_python: '>=3.9'
  official: true
  covers: WonderFence real-time guardrail evaluation of AI prompts and responses
  linked_from: https://docs.activefence.com/
- language: typescript
  registry: npm
  name: '@alice-io/wonderfence-ts-sdk'
  version: 2.1.0
  url: https://www.npmjs.com/package/@alice-io/wonderfence-ts-sdk
  install: npm install @alice-io/wonderfence-ts-sdk
  license: MIT
  published: '2026-08-03'
  official: true
  covers: TypeScript client for the WonderFence Trust & Safety evaluation API
- language: python
  registry: pypi
  name: parlant-alice
  version: 0.1.5
  url: https://pypi.org/project/parlant-alice/
  install: pip install parlant-alice
  repository: https://github.com/ActiveFence/parlant-alice
  official: true
  covers: >-
    Framework integration that wires the Alice SDK into the Parlant open-source AI agent
    framework; depends on wonderfence-sdk.
- language: typescript
  registry: npm
  name: '@alice-io/caterpillar'
  version: 1.0.11
  url: https://www.npmjs.com/package/@alice-io/caterpillar
  install: npm install -g @alice-io/caterpillar
  repository: https://github.com/alice-dot-io/caterpillar
  homepage: https://caterpillar.alice.io/
  license: MIT
  official: true
  bin: caterpillar
  covers: >-
    Open-source security scanner that inspects AI agent skill files (e.g. Claude Code skills) for
    dangerous or malicious behavior. Ships a CLI — see cli/activefence-cli.yml.
registries_probed:
- {registry: npm, scope: '@alice-io', result: found}
- {registry: npm, scope: '@activefence', result: none}
- {registry: pypi, names: [wonderfence-sdk, parlant-alice], result: found}
- {registry: pypi, names: [activefence, activefence-sdk, alice-io], result: none}
- {registry: maven-central, result: none}
- {registry: nuget, result: none}
- {registry: pkg.go.dev, result: none}
- {registry: rubygems, result: none}
- {registry: packagist, result: none}
- {registry: crates.io, result: none}