Back to directory

GoClaw

A multi-tenant AI agent gateway written in Go that rebuilds OpenClaw with multi-tenant PostgreSQL isolation, layered permission controls and native concurrency for deploying AI agent teams at scale.

GoClaw

GoClaw is a multi-agent AI gateway, described as OpenClaw rebuilt in Go, aimed at running multiple AI agent teams for many separate users or tenants without compromising on safety. It supports more than 20 LLM providers and seven messaging channels, including Telegram, Discord, Slack and WhatsApp, backed by multi-tenant PostgreSQL storage.

Its architecture includes an eight-stage agent pipeline from context gathering through thinking, acting, observing and summarizing; a four-mode prompt system with section gating and cache-boundary optimization; three-tier working, episodic and semantic memory with progressive loading; a document-based knowledge vault with wikilink-style links and hybrid full-text and vector search; agent teams with shared task boards and synchronous or asynchronous delegation; and a five-layer permission system with rate limiting, prompt-injection detection, SSRF protection and AES-256-GCM encryption for per-user API keys.

It compiles to a single, roughly 25 MB static Go binary and requires PostgreSQL 18 with pgvector.

More in Agent frameworks

See all 902
  • Recurse

    recurse.run

    A serverless harness that lets a coding agent design, test, and deploy its own specialist sub-agents as tools, MCP servers, or bots, defined by a schema-validated manifest.

  • ProofForge

    shulgin.is-a.dev

    An AI-agent pipeline that decomposes a mathematics problem, proves each piece, and formalizes the result in Lean 4 and Mathlib until the Lean kernel accepts it, with six of its proofs already merged into Google DeepMind's formal-conjectures repository.

  • local-coder

    github.com

    Builds a local coding-agent team on top of Ollama and OpenCode by detecting the machine's hardware, recommending models for seven specialist roles, and testing whether each assigned model can actually use tools, edit a repository and pass verification.