OpenViSTAI · Agents · Blockchain · Crypto
Submit a project
Menu
Agent Frameworks

Go Agent

Go Agent is an Apache-2.0 Go framework for building AI agents, with provider adapters, memory, file context, tool orchestration, and multi-agent coordination for applications.

Explore the practical guide ↓

Independent OpenViST project profile, based on linked sources.

Go Agent — OpenViST project profile in Agent Frameworks.
OpenViST project illustration. See the linked sources for product details.
Understand the project

Practical overview

When it fits

Use it when implementing agent behavior in a Go application and you want to compose the model, memory, tools, and coordination in code. The dummy model supports local wiring examples; a real provider requires its documented credentials.

Who it is for

  • Go developers adding AI agents to applications or services.
  • Teams assembling agents that need memory, tool access, or multi-agent workflows.
  • Developers building coding agents that retrieve focused context from Go repositories.

Documented uses

  • Create an agent that generates responses using a selected LLM provider.
  • Maintain session memory or configure vector-store-backed long-term memory.
  • Connect agents to UTCP tools, files, and specialist agents.
  • Build graph-based workflows with routing, fan-out, and join behavior.
  • Index Go code and assemble bounded source context for coding tasks.
Before you choose

Requirements & limits

The public GitHub repository lists an Apache-2.0 license. The README targets Go 1.25.10. Real model providers require provider-specific credentials; the documented dummy-model quick start does not require API keys.

  • The framework targets Go 1.25.10, so check that target against the Go environment for the intended project.
  • Choose the direct agent API for small programs and tests; the ADK is documented for reusable modules, shared sessions, provider selection, and UTCP runtime wiring.
  • Provider requirements differ: for example, OpenAI accepts OPENAI_API_KEY or OPENAI_KEY, while Vertex AI needs project and location settings plus Application Default Credentials.
  • Embedding configuration is separate from model selection; the README documents memory.AutoEmbedder and a DummyEmbedder fallback.
  • Graph workflows support explicit nodes and edges; join nodes wait for outputs from direct predecessors and can be bounded with a join timeout.

Check current requirements in the official documentation. Source-based guidance does not mean OpenViST has run or independently tested this project.

Source-based practical guide

Set up a local agent wiring example

Install the Go module and prepare the documented dummy-model example for local wiring.

Before you start

  • Go 1.25.10, the module's stated target version.
  • A Go project that can import github.com/Protocol-Lattice/go-agent.

Steps

  1. Add the module with the command below.

    Documented command
    go get github.com/Protocol-Lattice/go-agent

  2. Create an agent using models.NewDummyLLM and in-memory session storage, following the README quick-start example.

  3. Call Generate with a session ID and prompt, then handle the returned output and error.

Expected outcome

A small Go program is configured to generate a response through the dummy model and in-memory storage, without provider API keys.

Guide limitations

  • The dummy model and in-memory storage example is for local wiring and tests; it does not connect to a real LLM provider.
  • This guide covers setup and the documented example, not independently verified execution.
Documentation 1 ↗
Source-based practical guide

Configure an OpenAI-backed agent model

Select the documented OpenAI provider for an agent application.

Before you start

  • Go 1.25.10, the module's stated target version.
  • An OpenAI API key available as OPENAI_API_KEY or OPENAI_KEY.

Steps

  1. Add the Go Agent module to the Go project.

  2. Make the OpenAI credential available through one of the documented environment variables.

  3. Use models.NewLLMProvider with the openai provider name and a model identifier, as shown in the README.

  4. Pass the resulting model to the agent or ADK setup used by the application.

Expected outcome

The application is configured to select an OpenAI-backed model through Go Agent's provider adapter.

Guide limitations

  • The guide does not specify provider pricing or guarantee model availability.
  • The README's example uses gpt-4o-mini; model access depends on the provider credentials and account.
Documentation 1 ↗
COMPARE OPTIONS

Compare options for Agent orchestration

These projects document a common task and share a project role. This is a source-based comparison, not an execution benchmark or a guarantee of interchangeability.

ProjectDocumented purposePlatformAccess and requirements
Go AgentCreate an agent that generates responses using a selected LLM provider.Linux, macOS, Self-hosted, ServerThe public GitHub repository lists an Apache-2.0 license. The README targets Go 1.25.10. Real model providers require provider-specific credentials; the documented dummy-model quick start does not require API keys.
AgeniteDefine typed tools with JSON Schema validation and error handling.CLI
VeldraCreate an agent with a policy, tools, skills, knowledge base and reasoning method.Self-hostedThe documented deployment is self-hosted and uses a sign-in gate after a first-run administrator account is created.
Evidence and transparency

Sources & review

Official repository ↗
Source record details
Source availability
Active
Sources checked
12 October 2026
Original website title
Protocol-Lattice/go-agent · Official homepage title · retained in its source language
How we review sourcesSuggest a correction