HypercubsDRC Docs
tutorial developer public

Start here

Welcome to DRC. This guide takes you from your first visit to your first recorded run in about ten minutes. Work through the steps in order and stop when you have what you need.

DRC gives every execution an identity. When your software runs, most of what happened disappears with it. DRC keeps the run as an artifact you can open again: replay it exactly, compare two runs to find where they first differ, and prove an outcome with evidence. The way Git gives code a history you can return to, DRC gives every execution a history you can return to.

Understand the loop

The whole product is one loop. Keep it in view while you read the rest of the guide:


a run happens
  → DRC records it
  → you replay it exactly as it ran
  → you compare runs and find the first difference
  → you verify a change or build on what you learned
  • Record. A capture keeps what actually happened: the sequence of events, their order, and the side effects, not just a log line or a stack trace.
  • Replay. A recorded run plays back under controlled time, randomness, network, and side-effect policy, so it behaves the same way twice.
  • Compare. DRC diffs two executions and points at the earliest divergence, not just the final error.
  • Build. A fix, a deployment decision, or a claim about behavior can be tested against the recorded run and sealed into signed evidence.

You do not need to understand the machinery to start. The tutorial does that work for you.

Create your account

Start at the signup page. You can create an account with an email and password, or sign up with GitHub, Google, or your organization's SSO connection.

  1. Open the signup page and choose how you want to sign up.
  2. If you used email, check your inbox for the verification message and open the link it contains. The link expires after 60 minutes; request a new one from the sign-in screen if it has lapsed.
  3. Return to the sign-in screen and sign in.
  4. Open your profile and confirm your email shows as verified.

Result: your email is verified and you can sign in. You land in your first workspace, which starts as yours alone: a workspace can be isolated or shared, and you can invite other people into it whenever you are ready to collaborate.

For a production account, enable two-factor authentication from your profile security settings before you record anything you care about.

Land in your first workspace

A workspace is a shared, isolated place for one team's executions and evidence. What happens in your workspace stays inside its boundary, and only people you invite can see it.

A free account includes one hosted workspace, so one is ready for you after sign-in.

  1. From the workspace list, open your workspace.
  2. Look at the overview to see what is recorded there.
  3. Open the sections you will use most: executions, where recorded runs are listed, and evidence packs, where signed results are stored.

Result: you can move between the workspace areas and know what each one holds. If you are evaluating DRC alone, do not create an organization yet. Organizations are for teams and require a paid plan.

Record your first run

A run is one execution of your application, service, or workflow from start to finish. This is the moment the loop becomes real, so keep the first run small and non-sensitive.

  1. Choose something you can run from your own machine or network: a service endpoint, a locally hosted system, or a distributed system with a few moving parts.
  2. Capture the run. The capture records the execution and uploads it to your workspace under a name you choose.
  3. Open the executions list in your workspace and confirm the run is there.
  4. Open the recorded run and replay it. Replay runs it back under controlled conditions, so you can step through what happened without touching the original environment.
  5. If you have a second run that differs, compare the two and let DRC point at the first divergence.

Result: you have replayed a real run and can see where one execution differs from another. From here, replaying a failure is the same action you just performed on a success.

Decide what is next

  • Stay on the free plan and keep recording runs, replaying them, and comparing versions as you change code.
  • Bring a team. Buy a Team plan, create your organization during checkout, and invite teammates by email. Assign roles, then set up SSO from organization settings so your team signs in with your identity provider (Set up SSO for your organization). Hosted Continuous Verification, evidence packs, and Continuous Compliance are Team plan capabilities. Local Continuous Verification is available on Free. Start with Create an organization and invite your team.
  • Automate DRC. Use the CLI reference or HTTP API reference when you want captures, replays, and evidence to run inside CI and other automation you already have. Configure Continuous Verification and Continuous Compliance in [.drc/signal.yml](signal-yml.md).
  • Understand the boundaries. Read how retention and deletion work before you record production traffic, and confirm the capture policy redacts what your team does not want stored.

Next page: Capture and replay an execution. After you can capture and replay, configure Continuous Verification and Continuous Compliance in [.drc/signal.yml](signal-yml.md).