Lobu

Executive Summary

Lobu is multi-tenant, sandboxed agent orchestration for teams — "multi-tenant OpenClaw for organizations." It runs OpenClaw's Pi Agent runtime behind a hardened gateway with MCP proxy, multi-provider auth (16 LLM providers), and per-context isolation. Originally launched as peerbot.ai in July 2025, it has since consolidated on the OpenClaw runtime, relicensed from BUSL-1.1 to Apache-2.0, and launched a managed cloud (Lobu Cloud / Serverless OpenClaw). As of June 2026 the project is highly active: daily commits, v11.0.0 released May 30, 2026, and stars up from 36 to 162.

---

Product Overview

Lobu's core value: run powerful OpenClaw agents without exposing your infrastructure. All worker traffic routes through a gateway proxy — workers have no direct network access. Secrets stay isolated, MCP tools are proxied with scoped tokens, and network egress is domain-filtered. Lobu rewrites only the gateway layer (~40k LOC) to be multi-tenant while keeping OpenClaw's Pi harness untouched inside each worker.

Key Capabilities

Interfaces

---

Technical Architecture

Slack/Telegram/WhatsApp/Discord/Teams/API → Gateway → Worker (OpenClaw Pi Agent)
                                              ↑              ↓
                                     Postgres (pgvector)  HTTP/MCP Proxy
                                                              ↓
                                                   Domain-filtered Internet

Key Technical Details

---

Lobu vs OpenClaw

---

Strengths

• True isolation — Workers have no direct network access. All traffic routes through gateway with domain filtering.
• Secrets never exposed — MCP OAuth, provider credentials, and env vars resolved at gateway level. Workers never see client secrets.
• Genuinely open source — Relicensed from BUSL-1.1 to Apache-2.0, removing the commercial-use objection.
• Lightweight embedded mode — just-bash + Nix virtual filesystems at ~50MB per instance; 300 concurrent agents tested on a single machine without Docker.
• OpenClaw compatibility — Skills from ClawHub, IDENTITY.md, SOUL.md, USER.md all work. Migration path from vanilla OpenClaw.
• Enterprise deployment — Kubernetes via public Helm chart with NetworkPolicies, RBAC, gVisor/Kata. Production-ready.
• Shipping velocity — v9 through v11 released in May 2026 alone; commits landing daily as of June 2026.

---

Cautions

• Still early stage — 162 stars as of June 2026 (up from 36 in February, but tiny). Limited community validation; no independent reviews or discussion threads found.
• Fewer platforms — Slack, Telegram, WhatsApp, Discord, Teams, API. OpenClaw has 15+ integrations.
• Complexity — Gateway proxy architecture adds operational overhead vs simple single-node OpenClaw, though embedded single-process mode narrows the gap.
• Effectively solo maintainer — @buremba authored ~90% of commits; remaining contributions are small. Bus factor concern for enterprise adoption.
• Dropped dual runtime — The Claude Code SDK runtime option has been de-emphasized; the project now positions itself purely as multi-tenant OpenClaw.

---

Pricing & Licensing

Licensing model: Apache-2.0 (relicensed from BUSL-1.1; same codebase self-hosted and cloud, no vendor lock-in)

Hidden costs: Infrastructure if self-hosting (K8s cluster, VM, or Docker host), LLM API costs

Funding: Not publicly disclosed.

---

Competitive Positioning

Direct Competitors

When to Choose Lobu Over Alternatives

• Choose Lobu when: You need multi-tenant isolation for OpenClaw agents, self-hosted or serverless
• Choose Tensol when: You want managed hosting without infrastructure
• Choose OpenClaw when: You need maximum platform integrations (15+)
• Choose NanoClaw when: You want simpler container isolation

---

Ideal Customer Profile

Best fit:

• Teams deploying AI agents to Slack/Telegram with security requirements
• Organizations needing per-channel/user isolation
• DevOps teams comfortable with Kubernetes
• Companies wanting Claude Code + OpenClaw flexibility

Poor fit:

• Individual users wanting simple setup
• Those needing 15+ chat platform integrations
• Organizations requiring vendor-backed enterprise support
• Teams without infrastructure expertise (though embedded single-process mode lowers the bar)

---

Viability Assessment

Early-stage project with compelling architecture and visibly accelerating execution. The security-first approach differentiates from vanilla OpenClaw, and the Apache-2.0 relicense plus serverless cloud give it both an adoption and a revenue story. Still effectively a one-person project.

---

What Developers Say

No independent developer commentary on Lobu was found as of June 2026 — searches of Hacker News, Reddit, and the broader web surfaced only the project's own materials. At 162 stars, the project has not yet generated public community discussion. This section will be updated as third-party reviews appear.

---

Bottom Line

Lobu is "OpenClaw for teams" — adding the multi-tenant isolation, network sandboxing, and MCP proxy that enterprises need. Four months on, the project has clearly accelerated rather than stalled: relicensed to Apache-2.0, launched a usage-billed serverless cloud, added Discord and Teams channels, shipped v11, and grew from 36 to 162 stars. But it remains an effectively solo-maintained bet with zero independent community validation.

Recommended for: Teams deploying agents to Slack/Telegram/Discord/Teams who need isolation, self-hosted (single Node process or Kubernetes) or via Lobu Cloud

Not recommended for: Individual users, those needing many chat platforms, or organizations requiring vendor-backed enterprise support

Outlook: The trajectory since February 2026 — Apache-2.0, managed cloud, rapid releases — is exactly what an early infrastructure project should look like. If star growth converts into contributors and paying cloud users, Lobu could become the de facto "enterprise OpenClaw" layer. The security architecture is what OpenClaw should have built.

---

Research by Ry Walker Research • methodology