CARBON — Engineering ConfidenceUser Guide
CHAPTER 10 / 60 Core + Pro

/carbon-help

Analyze the current project, current change, existing tests, UI, API, and AI surfaces, CARBON state, confidence state, automation, and risk signals, then recommend the best CARBON commands and prioritized next steps.

Jay · Your AI test managerStart & investigate

When to use it

You are unsure which workflow matches the problem.

Before you start

Explain the symptom and the evidence you already have. Look for a small, justified next step; recommendations themselves are not executed tests.

Try this in your coding agent.

/carbon-help Which command should I use to investigate our new search feature?

Replace sample paths with your project. Slash names are shorthand; your host may use a plugin prefix or a skill picker. Invocation help →

ILLUSTRATIVE WALKTHROUGH

Search-feature investigation

An expanded teaching example, not a recorded or live execution. The report stays visible while the replay explains how to read it.

  1. 01Read context
  2. 02Analyze
  3. 03Show evidence
  4. 04Report

Coding-agent conversation

Read context

I’m reviewing search-feature investigation. I’ll keep the scope and permissions explicit.

Analyze

Analyze: Which command should I use to investigate our new search feature?

Show evidence

Here is the evidence trail to inspect: Question: missing results or confusing UI? Suggested commands: /carbon-issues, /carbon-api.

Report

Jay recommends an exploratory pass before focused API checks. A recommendation is not an executed check.

CARBON

Example report

Illustration
Query → catalog → visible result
Query → catalog → visible result — TestBucks teaching screen

TestBucks course screen with illustrative analysis. The screenshot supplies context; the demo does not establish a newly observed bug.

  1. Query intent
  2. Catalog / processing
  3. Visible result

Illustrated outcome

Jay recommends an exploratory pass before focused API checks.

Evidence to inspect

Question: missing results or confusing UI? Suggested commands: /carbon-issues, /carbon-api.

Still untested / not established

A recommendation is not an executed check.

Ready. Read the full example or play its steps.

What a useful result contains

A bounded claim

For search-feature investigation, state the target, revision, environment and actual work performed. Keep planned, executed, blocked and untested work separate.

An inspectable trail

Question: missing results or confusing UI? Suggested commands: /carbon-issues, /carbon-api. In a real run, retain the source, trace, screenshot or fixture behind the conclusion.

The qualification

A recommendation is not an executed check.

The next decision

Decide whether the evidence calls for a focused check, a reviewed fix or more context. Do not convert a plausible example into a verified finding.

Adapt the example to your project

  1. Replace search-feature investigation with a real feature, path or decision in your project.
  2. State the expected rule and its source. Include safe fixtures, constraints and a budget where appropriate.
  3. Ask for exact evidence and remaining gaps. Approve consequential actions separately.
/jay Review the /carbon-help result. Separate observed evidence, assumptions and untested scope, then recommend the next useful check.

Where to go next

/carbon →/carbon-test →

Content checked against the local command definition: skills/carbon-help/SKILL.md. Examples adapted from the homepage’s TestBucks walkthroughs. Availability is edition- and host-dependent.