Free & ProDifferent perspectives. One harness.
AI-generated testing perspectives, not human reviewers. Select a profile to explore its focus.
AIAI QA sub-agent
AI test manager
Modeled on Jason Arbon’s approach to testing: start with the customer’s intent, follow the consequential risks, challenge the evidence, and make the next useful action clear. Jay is an AI persona inspired by Jason Arbon, not Jason himself. His visual identity is a distinctive J-shaped character, not a human portrait.
A perspective available to CARBON—not a claim that this tester has reviewed your project. Findings require execution evidence.
AIAI QA sub-agent
Security sub-agent
Web-application security specialist focused on authentication, authorization, session safety, exposed data, dependencies, and practical vulnerability evidence.
A perspective available to CARBON—not a claim that this tester has reviewed your project. Findings require execution evidence.
AIAI QA sub-agent
OWASP sub-agent
OWASP-focused security specialist who challenges access control, injection, insecure design, misconfiguration, vulnerable components, integrity, logging, and SSRF risks.
A perspective available to CARBON—not a claim that this tester has reviewed your project. Findings require execution evidence.
AIAI QA sub-agent
Accessibility sub-agent
Accessibility specialist focused on keyboard use, assistive technology, inclusive interactions, responsive access, labels, feedback, and barriers affecting disabled users.
A perspective available to CARBON—not a claim that this tester has reviewed your project. Findings require execution evidence.
AIAI QA sub-agent
WCAG sub-agent
WCAG specialist focused on criterion-level evidence across perceivable, operable, understandable, and robust behavior, without overstating automated scan results as conformance.
A perspective available to CARBON—not a claim that this tester has reviewed your project. Findings require execution evidence.
AIAI QA sub-agent
Usability sub-agent
User-centered design and usability specialist focused on discoverability, task clarity, responsive behavior, interaction friction, visual hierarchy, and inclusive UX.
A perspective available to CARBON—not a claim that this tester has reviewed your project. Findings require execution evidence.
AIAI QA sub-agent
Compatibility sub-agent
Compatibility specialist focused on browser, device, viewport, operating-system, assistive-technology, and support-matrix evidence, with explicit coverage gaps rather than assumed portability.
A perspective available to CARBON—not a claim that this tester has reviewed your project. Findings require execution evidence.
AIAI QA sub-agent
Performance sub-agent
Performance specialist focused on user-visible latency, Core Web Vitals, payload and request cost, caching, API timing, scalability, mobile constraints, memory, and leaks.
A perspective available to CARBON—not a claim that this tester has reviewed your project. Findings require execution evidence.
AIAI QA sub-agent
Content sub-agent
Content and UX-writing specialist focused on page identity, clear copy, information architecture, credibility, navigation, status communication, readability, and content quality.
A perspective available to CARBON—not a claim that this tester has reviewed your project. Findings require execution evidence.
AIAI QA sub-agent
Localization sub-agent
Localization specialist focused on translation accuracy, locale formats, pluralization, truncation, bidirectional layout, cultural fit, regulatory copy, and recovery behavior across languages.
A perspective available to CARBON—not a claim that this tester has reviewed your project. Findings require execution evidence.
AIAI QA sub-agent
Data integrity sub-agent
Data-integrity specialist focused on agreement between rendered values, APIs, events, schemas, exports, storage, calculations, retention, and downstream state.
A perspective available to CARBON—not a claim that this tester has reviewed your project. Findings require execution evidence.
AIAI QA sub-agent
Forms sub-agent
Forms specialist focused on input contracts, validation, boundaries, state, submission, recovery, data quality, conversion barriers, and accessible interaction.
A perspective available to CARBON—not a claim that this tester has reviewed your project. Findings require execution evidence.
AIAI QA sub-agent
Error UX sub-agent
Error and recovery specialist focused on truthful status, clear messages, consistent severity, actionable recovery, safe logs, failure loops, and user trust.
A perspective available to CARBON—not a claim that this tester has reviewed your project. Findings require execution evidence.
AIAI QA sub-agent
Search sub-agent
Search-interface specialist focused on discoverability, query handling, suggestions, empty states, input accessibility, relevance feedback, and resilient result navigation.
A perspective available to CARBON—not a claim that this tester has reviewed your project. Findings require execution evidence.
AIAI QA sub-agent
News & content sub-agent
Digital publishing specialist focused on article identity, readability, supporting media, metadata, context, advertising boundaries, and multi-device content behavior.
A perspective available to CARBON—not a claim that this tester has reviewed your project. Findings require execution evidence.
AIAI QA sub-agent
Landing pages sub-agent
First-impression and conversion specialist focused on value clarity, trust, navigation, calls to action, responsive composition, dead ends, and page credibility.
A perspective available to CARBON—not a claim that this tester has reviewed your project. Findings require execution evidence.
AIAI QA sub-agent
Checkout sub-agent
Checkout and payment specialist focused on order accuracy, address and payment input, trust, retry safety, pricing truth, completion, and conversion-blocking failures.
A perspective available to CARBON—not a claim that this tester has reviewed your project. Findings require execution evidence.
AIAI QA sub-agent
Shopping cart sub-agent
Shopping-cart specialist focused on line-item state, quantities, promotions, totals, inventory changes, persistence, accessibility, and safe transition to checkout.
A perspective available to CARBON—not a claim that this tester has reviewed your project. Findings require execution evidence.
AIAI QA sub-agent
Pricing sub-agent
Pricing and subscription specialist focused on plan clarity, comparison, currency and locale, hidden conditions, billing cadence, conversion paths, and truthful claims.
A perspective available to CARBON—not a claim that this tester has reviewed your project. Findings require execution evidence.
AIAI QA sub-agent
Privacy sub-agent
Privacy specialist focused on personal-data exposure, consent, tracking, collection minimization, storage and transfer, debug leakage, user expectations, and AI-specific privacy risks.
A perspective available to CARBON—not a claim that this tester has reviewed your project. Findings require execution evidence.
AIAI QA sub-agent
GDPR sub-agent
GDPR specialist focused on lawful consent, personal-data exposure, international transfer, minimization, retention, security, data-subject rights, transparency, and accountability.
A perspective available to CARBON—not a claim that this tester has reviewed your project. Findings require execution evidence.
AIAI QA sub-agent
Cookie consent sub-agent
Cookie-consent specialist focused on prior consent, purpose and vendor choices, reject symmetry, preference persistence, withdrawal, tracking verification, and regional behavior.
A perspective available to CARBON—not a claim that this tester has reviewed your project. Findings require execution evidence.
AIAI QA sub-agent
Legal policy sub-agent
Policy-content specialist focused on discoverability, internal consistency, versioning, contact details, user rights, actual-product alignment, accessibility, and credible presentation.
A perspective available to CARBON—not a claim that this tester has reviewed your project. Findings require execution evidence.
AIAI QA sub-agent
AI-generated code sub-agent
AI-generated-code specialist focused on logic, boundaries, null and empty states, failure handling, API use, security, privacy, tests, code smells, state, and misleading AI shortcuts.
A perspective available to CARBON—not a claim that this tester has reviewed your project. Findings require execution evidence.
AIAI QA sub-agent
AI chatbot sub-agent
Conversational-AI specialist focused on task success, conversation quality, state, memory, recovery, privacy, safety, injection, tools, latency, integration, and nondeterminism.
A perspective available to CARBON—not a claim that this tester has reviewed your project. Findings require execution evidence.
In Claude Code, Codex, and other coding agents, default CARBON tests use your agent’s configured model access. Signed in with an eligible subscription? CARBON uses that allowance. Using API credentials? API billing applies.
Local storage ≠ local-only AI. Local runs keep files, evidence, and reports in your workspace. Your coding agent may send code, screenshots, and other context to its model provider.
Plan limits and extra-usage charges still apply. CARBON Pro licensing and CARBON Cloud are separate. Optional direct-model checks and external integrations can have separate costs and data flows. Command-usage analytics vary by edition—review its privacy policy and settings. Local does not mean offline or air-gapped.
Check /status in Claude Code or codex login status in Codex to confirm your billing route. A configured API key can select API billing instead of subscription usage. A standalone MCP client supplies the model access and execution tools; the MCP server does not include a subscription.
Provider documentation: Claude Code billing · Codex sign-in and billing.
Free for internal workplace use. No signup required. Uses your agent’s existing model access; its plan limits or API charges apply.
Install CARBON Light from Claude’s plugin directory. No ZIP upload is needed for this route.
Start a new Cowork task or Claude Code session, then run:
/carbon:carbonThis is /carbon with Claude’s plugin prefix. CARBON tests your current project when the target is clear. Otherwise, it asks what to test: share a URL, folder, repository, API, app, feature, or requirements. It runs a broad assessment within your testing budget and shows findings, evidence-qualified confidence, coverage gaps, and prioritized next steps. If the command is missing, check that CARBON is enabled, reload or start a new session, and look for it in the slash-command picker.
In builds with the Jay entrypoints, use /carbon:jay or /carbon:j in Claude to talk directly to Jay, your AI Test Manager Agent. Other hosts may show /jay, /j, or a plugin-prefixed skill. Ask a question, review results, or tell him what to test. He replies in a frank, practical voice inspired by Jason Arbon. He is an AI persona, not Jason. Questions do not automatically start tests. If these commands are absent, update the plugin or use /carbon:carbon.
Local execution requires a supported host and Node.js 22+; demo creation also needs Python 3.10+. Ordinary Claude chat does not run the local MCP server. Your organization may restrict plugins or local servers. Your Claude subscription or usage limits still apply.
If CARBON is not listed, check your Claude version, account and organization’s plugin permissions. Do not install an unrelated similarly named plugin.
For an explicitly supported local setup, download the manual Light package. In Cowork, use the custom plugin upload option. In Claude Code, extract it and run claude --plugin-dir /absolute/path/to/carbon. Use one CARBON edition at a time.
Free testing workflows with a portable visual Studio workspace. Findings, screenshot maps, and evidence-qualified confidence, inside your development workflow.
Download Studio Lite for CodexVersion 0.1.7 · MIT open source · No CARBON account or MCP server required.
/carbon. Choose CARBON Lite’s command in the skill picker if your host requires a prefix.Run /carbon-studio to open the portable HTML workspace. If you have no project open, give Jay a URL, folder, API, or feature to test.
CARBON Studio Pro adds the native local-MCP workspace, deeper specialist commands, continuous testing and team workflows. Pro, Team, and Enterprise customers receive the Pro package through their customer download link.
Compare paid plans · Already a customer? Get your download link
Lite uses portable HTML reports; it does not have Pro’s native live MCP interface. Ordinary ChatGPT web chat does not install this Codex plugin. Your coding agent’s model access and usage limits still apply.
Extract the ZIP. In Cursor’s MCP settings, add a local server with command node and argument /absolute/path/to/carbon/server/index.mjs. Add the bundled skills/ using your host’s skill mechanism.
Use the CARBON prompt/tools. Slash-command and embedded-preview support depend on your Cursor version.
Install CARBON Light from Claude’s plugin directory. No ZIP upload is needed for this route.
/plugin.Start a new Cowork task or Claude Code session, then run:
/carbon:carbonThis is /carbon with Claude’s plugin prefix. CARBON tests your current project when the target is clear. Otherwise, it asks what to test: share a URL, folder, repository, API, app, feature, or requirements. It runs a broad assessment within your testing budget and shows findings, evidence-qualified confidence, coverage gaps, and prioritized next steps. If the command is missing, check that CARBON is enabled, reload or start a new session, and look for it in the slash-command picker.
In builds with the Jay entrypoints, use /carbon:jay or /carbon:j in Claude to talk directly to Jay, your AI Test Manager Agent. Other hosts may show /jay, /j, or a plugin-prefixed skill. Ask a question, review results, or tell him what to test. He replies in a frank, practical voice inspired by Jason Arbon. He is an AI persona, not Jason. Questions do not automatically start tests. If these commands are absent, update the plugin or use /carbon:carbon.
Local execution requires a supported host and Node.js 22+; demo creation also needs Python 3.10+. Ordinary Claude chat does not run the local MCP server. Your organization may restrict plugins or local servers. Your Claude subscription or usage limits still apply.
If CARBON is not listed, check your Claude version, account and organization’s plugin permissions. Do not install an unrelated similarly named plugin.
For an explicitly supported local setup, download the manual Light package. In Cowork, use the custom plugin upload option. In Claude Code, extract it and run claude --plugin-dir /absolute/path/to/carbon. Use one CARBON edition at a time.
Extract the ZIP and configure its local MCP server: command node, argument /absolute/path/to/carbon/server/index.mjs. Use the included skills where your host supports them. Exact UI integration varies by host version.
{"mcpServers":{"carbon":{"command":"node","args":["/absolute/path/to/carbon/server/index.mjs"]}}}The server uses stdio MCP. It exposes eight prompts and six evidence/configuration tools. Your agent supplies browser, shell and model execution.
Install one edition at a time. Keep your project’s .carbon evidence when upgrading.
INSIDE YOUR AI CODING AGENT
Real testing, real evidence. Choose a useful assessment today—or an ongoing quality system for your team.
Free · no account required
A complete, bounded assessment inside the coding agent you already use.
Use it at work. No run quota or hidden findings.
Uses your agent’s existing model access; its plan limits or API charges apply.
Compare paid plans below
Continuous coverage, deeper investigations and your team in the workflow.
See current pricing and plan details. Questions before choosing? Email jason@testers.ai.
Codex Studio Lite is MIT open source. The Claude Light package uses PolyForm Perimeter, permitting internal commercial use but restricting competing products. Pro is commercially licensed. CARBON Cloud is a separate service.
Choose a plan for testing inside your AI coding agent. Compare the current prices and plan details below, then continue securely with Stripe.
Your coding agent’s model access and usage limits still apply. CARBON Cloud and CARBON Private Cloud are separate offerings, not included with these harness plans.
Install and sign into a supported coding agent first. CARBON adds testing instructions, specialist knowledge and evidence tools to that environment; it is not a separate model subscription.