AgentDock
Get started

Autonomous AI coding, under governance

Ship issues autonomously.
Stay on budget.

AgentDock orchestrates AI coding agents across GitHub and Azure DevOps - with preflight scoring, hard budget caps, and a full audit ledger. Turn engineering budgets into shipped pull requests.

No credit card required · Connect a repo in minutes

Issues enter the AgentDock engine, where preflight, testability scoring, quality and risk scoring, cost forecasting, budget gating, and queue ranking happen inside one system before build runs create pull requests.

How it works

From repo to merged pull requests in three steps.

  1. Connect your repo

    Add a GitHub or Azure DevOps repository, paste credentials, and AgentDock indexes your open issues - all secrets encrypted at rest with pgcrypto.

  2. Set a build wave

    Pick a spend ceiling, a concurrency limit, and a time window. AgentDock forecasts issue costs upfront and refuses to start runs that would breach the cap.

  3. Sleep. Wake to PRs.

    The dispatcher preflights, ranks, and dispatches issues with bounded concurrency. Every transition lands in the audit ledger - and merged pull requests land in your queue.

Why teams choose AgentDock

Outcomes you can measure on the first night.

Ship while you sleep

Queue a backlog at end of day, set a budget, and step away. You review merged pull requests over morning coffee - no babysitting, no overnight pager.

Hard budget caps

Set a ceiling and AgentDock treats it as law. No surprise invoices, no runaway agents - your AI spend stays as predictable as any other line item.

Cut wasted runs

Unclear or risky work is held back before a single line of code is written. Pay pennies to analyse a bad issue instead of dollars to fail at it.

Multi-provider

Run GitHub and Azure DevOps side by side under one budget model and one audit trail. Adopt AgentDock without re-platforming or locking yourself in.

What's inside

A control plane for AI-assisted software delivery.

Preflight analysis

Auto-score issues on clarity, risk, testability, and estimated cost before they reach the build queue.

Build waves with autonomous dispatch

Budget pools and time-bounded build waves drive a dispatcher that walks the backlog with bounded concurrency and clean stop conditions.

Dual-layer cost guardrails

Per-run ceilings plus per-build-wave caps. Runs cancel themselves if cost or elapsed time exceeds limits.

Smart issue ranking

Score candidates on six dimensions - quality, risk, conflict, age, cost, history - with configurable strategies and per-repo weights.

Encrypted credential storage

API keys and tokens live in Postgres with pgcrypto encryption. No plaintext secrets in .env.

Append-only audit ledger

Every invocation, dispatch decision, and skip reason is recorded with actor, timestamp, and outcome.

Run lifecycle tracking

From spawn through preflight, build, and cost attribution - every step is observable end-to-end in the dashboard.

Multi-provider support

Connect GitHub and Azure DevOps repositories in a single dashboard. The same preflight, build, and dispatch workflows apply across both platforms.

Ready to govern your AI coding agents?

Create your account and dispatch your first build wave in minutes — or explore how it works first.