CARBON — Engineering ConfidenceUser Guide
CHAPTER 34 / 60 Pro

/carbon-confidence

Run an intent-first, bounded confidence investigation loop for a product or release: prioritize consequential questions, investigate and challenge, reassess evidence, then reprioritize. Persist promises, evidence, human context and judgment ranges, with a live in-agent decision audit.

Jay · Your AI test managerProof & release confidence

When to use it

You have evidence but still need a reasoned shipping judgment.

Before you start

Start with intended outcomes and the costliest unanswered questions. Link each confidence judgment to fresh evidence, missing coverage and human assumptions; do not interpret it as a probability of having no defects.

Try this in your coding agent.

/carbon-confidence Assess confidence in shipping the checkout changes using the available evidence.

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

Checkout shipping confidence

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

  1. 01Read evidence
  2. 02Assess gaps
  3. 03Investigate
  4. 04Synthesize

Coding-agent conversation

Read evidence

I’m reviewing checkout shipping confidence. I’ll keep the scope and permissions explicit.

Assess gaps

Assess gaps: Assess confidence in shipping the checkout changes using the available evidence.

Investigate

Follow the weakest evidence link and record what remains blocked.

Synthesize

Evidence supports the happy path, but recovery prevents a strong shipping recommendation. A confidence opinion is not a probability of zero defects.

CARBON

Example report

Illustration
Follow the cart state
Follow the cart state — TestBucks teaching screen

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

  1. Cart input
  2. Totals & state
  3. Checkout evidence

Illustrated outcome

Evidence supports the happy path, but recovery prevents a strong shipping recommendation.

Evidence to inspect

Link requirements → checked paths → unresolved timeout risk → confidence rationale.

Still untested / not established

A confidence opinion is not a probability of zero defects.

Ready. Read the full example or play its steps.

What a useful result contains

A bounded claim

For checkout shipping confidence, state the target, revision, environment and actual work performed. Keep planned, executed, blocked and untested work separate.

An inspectable trail

Link requirements → checked paths → unresolved timeout risk → confidence rationale. In a real run, retain the source, trace, screenshot or fixture behind the conclusion.

The qualification

A confidence opinion is not a probability of zero defects.

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 checkout shipping confidence 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-confidence result. Separate observed evidence, assumptions and untested scope, then recommend the next useful check.

Where to go next

/carbon-human →/carbon-release →

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