DEV.co
RAG Frameworks · puppyone-ai

puppyone

Puppyone is a Git-native platform that hosts context (documents, data, configurations) for AI agents with version control and file-level permissions. It exposes a shared Context Drive through multiple access points (CLI, MCP, REST API, Git, sandbox environments) so different agents can collaborate on shared knowledge with fine-grained security.

Source: GitHub — github.com/puppyone-ai/puppyone
861
GitHub stars
71
Forks
TypeScript
Primary language
Apache-2.0
License (OSI-approved)

Key facts

Objective fields from the source. Values we can't verify are shown as “Unknown” rather than guessed.

FieldValue
Repositorypuppyone-ai/puppyone
Ownerpuppyone-ai
Primary languageTypeScript
LicenseApache-2.0 — OSI-approved
Stars861
Forks71
Open issues26
Latest releasev0.0.3 (2026-05-31)
Last updated2026-07-05
Sourcehttps://github.com/puppyone-ai/puppyone

What puppyone is

Built on FastAPI (Python 3.12+) backend, Next.js 15 frontend, and Supabase PostgreSQL with S3/MinIO storage. Provides file-level security enforcement, Git-style checkout/commit workflows, Docker and E2B sandbox isolation, and task queuing via ARQ (Redis). Deployable as cloud-hosted or self-hosted Docker stack.

Quickstart

Get the puppyone source

Clone the repository and explore it locally.

terminalbash
git clone https://github.com/puppyone-ai/puppyone.gitcd puppyone# follow the project's README for install & configuration

Need it deployed, integrated, or customized instead? DEV.co ships production installs.

Best use cases

Multi-agent AI workflows with shared knowledge

Puppyone excels when multiple AI agents (Cursor, Claude Desktop, custom services) need read/write access to overlapping documents with enforced permissions, versioning, and audit trails per agent.

Self-hosted AI deployment with compliance requirements

Organizations that need on-premises control can deploy the full stack via Docker Compose. File-level security, audit logs, and Git-like workflows support audit and compliance audits.

RAG context management for agentic systems

Puppyone's connectors (GitHub, Google Drive, web pages, Supabase) and agent-friendly output formats (Markdown, JSON) streamline building and updating RAG datasets without manual ETL.

Implementation considerations

  • Self-hosted Docker stack requires Docker daemon access on the host; for remote/multi-tenant deployments, the docs recommend E2B sandboxing instead of Docker socket sharing.
  • Cloud deployments require Anthropic API key for agent chat; OAuth connectors (GitHub, Gmail, Google Drive) need provider credentials configured separately in backend environment.
  • File-level security enforcement is built into the filesystem layer, but requires correct permission setup during agent provisioning; misconfiguration could expose unintended files.
  • Git-style checkout/commit workflow introduces potential for merge conflicts and locking; agents must handle commit failures gracefully.
  • MCP endpoint creation and API key distribution must follow your organization's secrets management practices; no automatic rotation or vault integration is mentioned.

When to avoid it — and what to weigh

  • Single-agent or human-centric workflows — If your primary use case is human document collaboration, conventional file-sharing tools (Notion, Confluence, Drive) will be simpler. Puppyone's agent-level auth and versioning add overhead when not needed.
  • Real-time collaborative editing for humans — Puppyone is designed for agent read/write patterns (checkout, commit, conflict resolution), not simultaneous human editing. No built-in CRDT or WebSocket-based live sync is documented.
  • Production use without vendor support or SLA — The project is young (created Dec 2024, v0.0.3 as of latest release). Self-hosting requires operational expertise; cloud offering maturity and support terms are not stated in available data.
  • Closed-source or proprietary deployments — Apache 2.0 requires source code disclosure for derivative works in network-accessible deployments in some interpretations. Internal legal review is recommended before wrapping Puppyone in a proprietary service.

License & commercial use

Licensed under Apache License 2.0. This is a permissive OSI license that permits use, modification, and distribution for any purpose (personal, commercial, open-source, proprietary), provided copyright notice and LICENSE file are preserved and significant modifications are documented.

Apache 2.0 permits commercial deployment and proprietary use. However, if Puppyone is modified and deployed as a network service, some interpretations of the AGPL-adjacent requirements (§5) may apply; legal review is recommended for SaaS offerings. Cloud-hosted Puppyone and self-hosted forks for internal use are unambiguously permitted. No commercial support, SLA, or warranty is stated in the available documentation.

DEV.co evaluation signals

Editorial assessment — not user reviews. Directional, with an explicit confidence level.

SignalAssessment
MaintenanceActive
DocumentationAdequate
License clarityClear
Deployment complexityModerate
DEV.co fitGood
Assessment confidenceMedium
Security considerations

File-level security is enforced at the filesystem layer; correctness depends on proper permission configuration. Local Docker Compose stack shares the host Docker daemon with the backend container, which carries sandbox escape risk; docs recommend E2B for remote/multi-tenant. Auth uses Supabase JWT + Access Keys; no mention of rate limiting, secret rotation, or audit log tamper-proofing. Self-hosted deployments inherit PostgreSQL, MinIO, and Redis security postures; hardening is operator responsibility. No penetration test or security audit results are disclosed.

Alternatives to consider

Langchain (LangSmith) or Anthropic Workbench

Simplified agent context management with less infrastructure overhead; better integration with specific model providers. Trade-off: less fine-grained file-level security and Git-style versioning.

Supabase + custom vector store (e.g., Pinecone, Weaviate)

If you only need RAG and context retrieval (not agent-facing file access), a database + vector store is simpler to deploy and maintain. Trade-off: no Git versioning, agent permissions, or multi-access-point abstractions.

Hugging Face Spaces or Modal for agent orchestration + cloud storage (S3, GCS)

Lightweight alternative for stateless agent workflows with minimal shared context. Trade-off: no file-level security, audit logging, or agent-centric permission model.

Software development agency

Build on puppyone with DEV.co software developers

Start with Puppyone Cloud (no infrastructure required) or self-host via Docker. Review the Apache 2.0 license and documentation to confirm fit for your use case.

Talk to DEV.co

Related open-source tools

Surfaced by semantic similarity across the DEV.co open-source index.

Related on DEV.co

Explore the category and the services that help you build with it.

puppyone FAQ

Can I run Puppyone in production on-premises?
Yes, Docker Compose deployment is supported for self-hosted. However, ops burden (PostgreSQL hardening, S3 setup, OAuth, Redis tuning, E2B sandbox integration) is on you. No managed support SLA is documented.
Does Puppyone support my existing auth system (SAML, Okta, etc.)?
Not directly. It uses Supabase Auth (JWT + Access Keys). For corporate SSO, you would need a custom auth proxy or Supabase Enterprise. Details are not in provided docs.
What happens if two agents try to edit the same file concurrently?
Puppyone uses a Git-style checkout/commit workflow with conflict detection and locking. The second agent's commit would fail; it must resolve conflicts and retry. Automatic merge is not mentioned.
Is there a cost to using the cloud version?
Pricing is not stated in the provided data. Visit puppyone.ai for current terms.

Work with a software development agency

From first prototype to production, DEV.co delivers software development services around tools like puppyone. Our software development agency staffs experienced software developers and web developers for custom software development, web development, integrations, and ongoing support across rag frameworks and beyond.

Ready to manage AI agent context at scale?

Start with Puppyone Cloud (no infrastructure required) or self-host via Docker. Review the Apache 2.0 license and documentation to confirm fit for your use case.