What is Agent Beacon used for?

Agent Beacon is an open-source, local-first memory layer that records AI coding agent sessions across tools and turns useful workflows into reusable knowledge. It is a good fit for sharing debugging knowledge between coding agents, auditing what AI agents did in a repository and keeping agent context when switching tools.

Is Agent Beacon open source?

Yes. Agent Beacon is open source under the MIT licence. Its source code is on GitHub at Asymptote-Labs/agent-beacon and is written mainly in Go.

Is Agent Beacon free?

Yes. Agent Beacon is open source, so the software itself is free to use. A managed cloud version is also available.

What are some alternatives to Agent Beacon?

Similar open-source tools in the Developer Tools category include Agentacct, Munder Difflin and Mission Control. SaaS products in the same category include 1DevTool, AgentKit and CNVS.

Is Agent Beacon actively maintained?

Yes. The most recent commit to Agent Beacon was on 29 September 2026, and the latest release is v1.3.29, published on 28 September 2026. The project has 1.7k stars on GitHub.

Agent Beacon: open-source developer tools

About Agent Beacon

Agent Beacon is a memory layer for AI coding agents. It captures session history from many agent harnesses, including Claude Code, Cursor, Codex, OpenCode and Cline, and keeps it in one place so that a problem solved by one agent does not have to be rediscovered by another.

From that history it identifies useful workflows, corrections, debugging patterns and repository conventions, and makes the reviewed knowledge available to future agents through MCP or Agent Skills. Sessions can be replayed exactly, with prompts, responses, tool calls, commands, edits, approvals and token usage in a single trace. Data is stored as durable JSONL with explicit destinations.

The project is written in Go and licensed under MIT, and it describes itself as local-first with no harness lock-in. Setup can connect to a Beacon Cloud option through beacon.sh, while system, package, MDM and CI installations stay noninteractive and account-free. Its topics also mention observability and security telemetry, so it may suit developers and engineering teams who want visibility into agent activity.

Key features

Session capture across 20+ agent harnesses
Reusable knowledge from workflows and corrections
MCP and Agent Skills retrieval
Exact replay of prompts, tools, and edits
Durable JSONL storage with explicit destinations
Noninteractive installs for CI and MDM

Good fit for

→Sharing debugging knowledge between coding agents
→Auditing what AI agents did in a repository
→Keeping agent context when switching tools

Built with: Go
Tags: ai-agents; agent-memory; mcp; claude-code; observability; telemetry; local-first; go; knowledge-base

Agent Beacon: questions and answers

What is Agent Beacon used for?: Agent Beacon is an open-source, local-first memory layer that records AI coding agent sessions across tools and turns useful workflows into reusable knowledge. It is a good fit for sharing debugging knowledge between coding agents, auditing what AI agents did in a repository and keeping agent context when switching tools.
Is Agent Beacon open source?: Yes. Agent Beacon is open source under the MIT licence. Its source code is on GitHub at Asymptote-Labs/agent-beacon and is written mainly in Go.
Is Agent Beacon free?: Yes. Agent Beacon is open source, so the software itself is free to use. A managed cloud version is also available.
What are some alternatives to Agent Beacon?: Similar open-source tools in the Developer Tools category include Agentacct, Munder Difflin and Mission Control. SaaS products in the same category include 1DevTool, AgentKit and CNVS.
Is Agent Beacon actively maintained?: Yes. The most recent commit to Agent Beacon was on 29 September 2026, and the latest release is v1.3.29, published on 28 September 2026. The project has 1.7k stars on GitHub.