From Beamt — Q2 2026 Early Access

Hand Talos a spec, get back shipped code

Your autonomous engineering crew.

Forge new software and document what it does in plain English — autonomously, day and night. Built for solo founders and small engineering teams. Powered by Claude (and any AI). Q2 2026 early access.

What is Talos?

Talos forges new software and documents what that software does in plain English — autonomously, using AI.

Drop a SPEC.md into your repo, and Talos breaks it into GitLab or GitHub issues. You move issues to To-Do (or set rules — “always auto-work bugs”), and Talos plans, codes, opens MRs, watches CI, self-heals on failure, and writes Gherkin scenarios that document what each feature actually does. Day and night. Following your rules.

Bug reports submitted by your users become To-Do items. Security findings from agent scans become To-Do items. Whatever lands in your issue tracker can be auto-worked if you choose — or held for manual review if you don’t. The board is yours; Talos just keeps it moving.

How Talos works

From spec to shipped — you stay in your issue tracker.

1. Drop a SPEC.md

Or open issues directly. Or let an agent scan for security issues and open them for you.

2. Set the work rules

Drag issues to To-Do manually, or auto-work bug reports, security findings, or user-submitted issues.

3. Talos executes

Plans the work, opens a clone of your repo in an ephemeral Beamt-hosted container, writes the code, opens an MR, watches CI, self-heals on failure.

4. Gherkin docs ship alongside the code

Every feature gets a generated Gherkin scenario file: a plain-English record of what was built and what passes. Living documentation, not stale comments.

What makes Talos different

Cursor, Copilot, Devin, and the rest play in different lanes. Here’s where Talos lives.

Issue-tracker-native, not chat-driven

Cursor, Copilot, and Claude Code wait for you to type. Talos pulls from GitLab or GitHub issues — the same backlog your team already manages.

Continuous, governed pipeline, not single-task

Devin and SWE-agent run once. Talos runs day and night against your To-Do column, following your auto-work rules.

Generates Gherkin docs alongside code

You don’t just get an MR — you get a plain-English record of what each feature does. Living documentation, not stale comments.

AI-agnostic

Starts with Claude; any model that can drive Talos’s agent loop will be supported. You’re not locked into one provider.

Who Talos is for

Solo founders shipping a product without a team. Engineering teams of two to ten who are tired of the backlog growing faster than the burndown. Anyone who already lives in GitLab or GitHub issues and wants the To-Do column to move on its own.

Talos is not a code-completion tool — Cursor, Copilot, and Claude Code already serve that need. It is not a one-shot agent — Devin and SWE-agent already serve that. Talos sits in the gap between: a continuous, governed, issue-driven pipeline that ships working code and the plain-English documentation that proves it works, day in and day out.

A typical day: you drop a SPEC.md describing a new checkout flow. Talos opens seven issues against your repo — three test scaffolds, two service-layer changes, one new API endpoint, one Gherkin scenario. You drag the first three to To-Do; the rest stay in Backlog until you’re ready. Talos picks up issue #1, plans the work, opens an MR with the scaffolding, watches your CI run, and posts back when the pipeline goes green. By morning, three MRs are queued for your review and the Gherkin scenarios have been written and cross-linked from each.

Security & data model

Built by a former MedTech CTO who took HIPAA and HITECH compliance seriously for two decades. That mindset carries over here.

  • Beamt-hosted infrastructure

    Talos doesn’t run on your servers.

  • Ephemeral Docker containers

    Every task gets a fresh container, deleted on completion.

  • Repo clones, not direct access

    Talos pulls a clone, pushes branches; never holds your tree.

  • Secrets stay in your CI

    GitLab and GitHub secrets are never copied into Talos’s environment.

  • Regulated-software discipline

    Same architectural pattern used in FDA-audited systems where data exposure has legal consequences.

Languages & AI engines at launch

At launch (Q2 2026)

  • Languages: Python, JavaScript / TypeScript (Node)
  • AI engines: Claude (Sonnet, Opus)

Coming later

  • Languages: Go, Rust, Java, more
  • AI engines: GPT, Gemini, others

Usage-based pricing

You pay for the work Talos does — not for seats, not for repo access, not for users on your team.

Early-access pricing details ship with the Q2 2026 launch. Get on the list to be first.

Built by someone who’s shipped regulated, mission-critical software

Talos is built by Greg Smethells, former CTO of Medstrat — an FDA-approved medical imaging platform that served 25%+ of US orthopedic practices and reached $10M ARR before its acquisition by Zimmer Biomet (Fortune 500). Twenty years of writing software that runs in operating rooms and passes FDA audits. Now applying that same discipline to autonomous software engineering.

More about Greg →

Common questions

What does Talos do?

Talos forges new software and documents what that software does in plain English — autonomously, using AI. You write a SPEC.md, set work rules, and Talos handles planning, coding, MR opening, CI watching, self-healing, and Gherkin documentation.

How does Talos differ from Cursor or Copilot?

Cursor and Copilot wait for you to type. Talos pulls from GitLab or GitHub issues — the same backlog your team already manages — and runs day and night against your To-Do column. It’s a governed pipeline, not an interactive editor.

What programming languages does Talos support at launch?

Python and JavaScript/TypeScript (Node). More languages — Go, Rust, Java, others — will follow after the Q2 2026 launch.

Which AI engines does Talos support?

Claude (Sonnet, Opus) at launch. Any AI engine that can drive Talos’s agent loop will follow. You won’t be locked into one provider.

Where does Talos run? Is my code safe?

On Beamt-hosted infrastructure, in ephemeral Docker containers that are deleted as soon as the task is complete. Talos works from a clone of your repo, never the original tree, and never holds onto secrets — those stay in your CI/CD.

Does Talos work with my existing CI/CD?

Yes. Talos opens MRs and pushes branches; your CI runs against them as it would for any human-authored MR. If CI fails, Talos reads the logs and pushes a follow-up commit.

How is Talos priced?

Usage-based. You pay for the work Talos does, not for seats, repo access, or users on your team. Early-access pricing details ship with the Q2 2026 launch.

Can Talos auto-fix bugs reported by my users?

Yes. Configure auto-work rules per issue type — for example, “always auto-work bug reports” — and Talos will pick them up the same way it picks up any To-Do item.

Talos goes live in Q2 2026.

Drop your email at the top of this page to be first in line.

Get early access