STATEPORT / FOR YOUR CODING AGENT

Current headline: OPEN THE RESULT.

Open the result. Review the choices. Reopen the case.

Work with your agent in your local app. Inspect the result and what was actually checked. When the setup matters, keep a case ready for the next change.

Terminal — your project

# Run in your project, or ask your agent.

Choose your shell

01 / LOOK PAST “DONE”

The code changed.
Did the result?

A fallback can give an exhausted account another 100 credits. Use StatePort to open the local result, inspect the choice and recheck the same conditions.

How the loop works ↗

ILLUSTRATIVE EXAMPLE / NOT AN EXECUTED RUN

Zero credits should stay zero.

Same role, inputs and API response. Quota: 0. Request: 10.

01 / OPEN THE LOCAL RESULT

Buggy result

100 available

Allocation enabled
Expected after fix

0 available

Allocation disabled
02 / Agent-reported choice
Missing isn’t zero. Default only a missing quota → allocation stays disabled. Bug source ↗ · Proposed fix ↗
03 / Recheck the same conditions
Keep this case for the next change. Not checked: other roles or browser versions.
Explore the example ↗

02 / YOUR WORK. IN VIEW.

Look closer. Then decide.

01 / TRY IT

Open & Verify

Open the result in your local app. Try the behaviour. Inspect what was checked and what still needs attention.

Open and verify the result ↗
02 / UNDERSTAND IT

Review the decisions.

See important choices reported by your agent, their source and consequences. Keep the trade-off or ask for a different approach.

Review decisions ↗
03 / REVISIT IT

Keep a reusable case.

When browser setup matters, save the supported conditions. Reopen the case after your next change, or share it with another developer.

Fix and recheck the case ↗

Start with source, a live app or a saved case. Source help does not require a captured case.

03 / IN YOUR CURRENT CHAT

One command.
Your next task.

  1. Connect. Run the command and review local permissions.
  2. Ask. Give your agent the local app address and what you want to inspect or change.
  3. Inspect. Open the result, review the choices, keep it or continue.
Start with your agent ↑

04 / ON YOUR MACHINE

Local work.
Your call.

Setup downloads a local runtime. Browser work uses a separate profile. Cases stay on your machine unless you share them; your coding agent may process context remotely.

Security and data flow ↗

Your project. Your coding agent.

Run StatePort in your project, choose the agents to connect, and continue with the work in your current chat.

Setup and troubleshooting ↗

Prefer to explore by hand?

Optional Desktop gives you a visual workspace for cases and run evidence. Same product. Same license.

05 / TRY IT ON YOUR WORK

Your next change.
A closer look.

Connect your agent.
Choose when to start the trial.

Compare all plans ↗

DEVELOPER TRIAL

Try the
whole loop.

  • Open & Verify on local code
  • Start explicitly after email verification
  • No payment card required
14 daysYOUR WORK.
YOUR CALL.
Start with your agent

Continue with Solo Monthly — $12/month, or Founding Solo — $99 once ↗