# Track (testing itself)

> **An AI-first system to manage a software project and test it.** Point it at an app. Track learns what the app does, proves it with tests that only count when t

Source: https://track.letmetestit.com/p/track

## Mission

**An AI-first system to manage a software project and test it.**

Point it at an app. Track learns what the app does, proves it with tests that only count when they replay and pass, tells you whether you can ship, asks people only what it cannot see for itself, and keeps all of it true as the app changes.

## The problem

**Software ships without anyone knowing what works, and the tools meant to tell us cost people time and trust.**

- Poor software quality costs the US at least **$2.41 trillion a year** (CISQ, Nov 2022: it-cisq.org/wp-content/uploads/sites/6/2022/11/CPSQ-Report-Nov-22-2.pdf).
- At Google about **16% of tests** show some flakiness and **84%** of pass-to-fail transitions involve a flaky test (Google Testing Blog, May 2016: testing.googleblog.com/2016/05/flaky-tests-at-google-and-how-we.html).
- Atlassian estimates flaky tests cost it about **150,000 developer hours a year** (Atlassian, Dec 2025: atlassian.com/blog/how-we-build/taming-test-flakiness-how-we-built-a-scalable-tool-to-detect-and-manage-flaky-tests).
- **60%** of quality-engineering leaders name AI hallucination or reliability as a barrier (World Quality Report 2025, OpenText / Capgemini / Sogeti, Nov 2025), and more developers distrust AI accuracy (**46%**) than trust it (**33%**) (Stack Overflow Developer Survey 2025: survey.stackoverflow.co/2025/ai).
- AI test automation is projected to grow from **$8.81B (2025) to $35.96B (2032)** (MarketsandMarkets: marketsandmarkets.com/PressReleases/ai-test-automation.asp).

**Where the market is today:** the leading AI-testing services make their tests trustworthy with people in the loop (QA Wolf's human investigation, Checksum's human engineer final verification) or meter AI use in credits (Momentic, Functionize, Reflect, Autify). None of them states that a test counts only when it replays and passes. Track does, and proves it on itself.

## One engine, two products

- **For companies (B2B):** for developers, QA, founders and boards. Track learns each app, tests it (screens and backend), manages the work, and gives a ship verdict anyone can read.
- **For anyone:** a personal agent that acts for you on any website, with your permission, under guardrails. The same senses that test an app can carry out tasks on the web, and every automation is proven by replay before you rely on it.

The knowledge and management layer is where people live; testing is the first thing its senses do.

## Where it stands

**Not ready to ship v1.0 Launch: 161 blockers.**

In the whole app: 40 of 221 features proven by tests that replay and pass. 165 of 241 tests passing, including 83 of 85 backend (API) checks.

In v1.0 Launch: 0 of 5 features proven.

Blocking the release:
- Runs priced and held before they start has no test yet
- Accounts and deals has no test yet
- Support cases has no test yet
- Campaign drafts and approval has no test yet
- [launch-blocker][security] W-894 audit failures before unwrap must retain denied/expired attempts
- [launch-blocker][security] W-894 activation must reject retired-version rollback
- [launch-blocker][security] Secrets acceptance needs a hard full-history/tree CI credential scanning gate
- [launch-blocker][security] Broker exact resource equality must survive Postgres JSONB key ordering
- [launch-blocker][security][high] Vault forwards provider errors named SecretsError with plaintext material
- [launch-blocker][security][high] Broker confirms revoked while reserved operation can dispatch afterward
- [launch-blocker][security][medium] Broker replay digest binds mutable body instead of dispatched bytes
- [launch-blocker][accessibility] Secrets dark empty-state text fails axe contrast through data Bind defaults
- [launch-blocker][security] Scan triage: generic-api-key in packages/databases/tests/neon.test.js
- [launch-blocker][security] Scan triage: stripe-access-token in packages/governance/tests/signoffs.test.js
- [launch-blocker][security] Scan triage: stripe-access-token in packages/governance/tests/exports.test.js
- [launch-blocker][security] Scan triage: generic-api-key in packages/billing/tests/fixtures/stripe-test-mode.json
- [launch-blocker][security] Scan triage: generic-api-key in packages/deploy/tests/governance.test.js
- [launch-blocker][security] Scan triage: generic-api-key in packages/deploy/tests/helpers.js
- [launch-blocker][security] Scan triage: stripe-access-token in packages/billing/tests/money-stripe-adapter.test.js
- [launch-blocker][security] Scan triage: generic-api-key in packages/site/tests/helpers/env.js
- [launch-blocker][security] Scan triage: generic-api-key in packages/site/tests/godaddy.test.js
- [launch-blocker][security] Scan triage: generic-api-key in packages/site/tests/domains.browser.test.js
- [launch-blocker][security] Scan triage: generic-api-key in packages/metering/tests/spend.test.js
- [launch-blocker][security] Scan triage: generic-api-key in packages/outreach/tests/helpers.js
- [launch-blocker][security] Scan triage: generic-api-key in packages/outreach/tests/races.thorough.test.js
- [launch-blocker][security] Scan triage: generic-api-key in packages/outreach/tests/shared-suppression.test.js
- [launch-blocker][security] Scan triage: generic-api-key in packages/site/tests/pages.browser.test.js
- [launch-blocker][security] Scan triage: generic-api-key in packages/app/tests/needs.test.js
- [launch-blocker][security] Scan triage: generic-api-key in packages/app/src/testing.js
- [launch-blocker][security] Scan triage: generic-api-key in packages/app/template/tests/app.test.js
- [launch-blocker][security] Scan triage: generic-api-key in packages/voice/pack/tests/routes.test.js
- [launch-blocker][security] Scan triage: generic-api-key in packages/ai/pack/tests/fixtures.js
- [launch-blocker][security] Scan triage: generic-api-key in packages/ai/pack/tests/route.test.js
- [launch-blocker][security] Scan triage: generic-api-key in packages/ai/pack/tests/tools.test.js
- [launch-blocker][security] Scan triage: generic-api-key in packages/server/tests/helpers.js
- [launch-blocker][security] Scan triage: generic-api-key in packages/data/tests/client.test.js
- [launch-blocker][security] Scan triage: stripe-access-token in packages/payouts/tests/pinned.test.js
- [launch-blocker][security] Scan triage: generic-api-key in packages/files/tests/pinned.test.js
- [launch-blocker][security] Scan triage: private-key in tools/pinned/fixtures/leaf.key
- [launch-blocker][security] Scan triage: generic-api-key in packages/voice/tests/thorough.test.js
- [launch-blocker][security] Scan triage: generic-api-key in packages/ai/tests/thorough.test.js
- [launch-blocker][security] Scan triage: stripe-access-token in packages/payouts/tests/races.thorough.test.js
- [launch-blocker][security] Scan triage: generic-api-key in packages/mcp/tests/upgrade.test.js
- [launch-blocker][security] Scan triage: generic-api-key in packages/integrations/tests/thorough.test.js
- [launch-blocker][security] Scan triage: stripe-access-token in packages/ui/tests/states.browser.test.js
- [launch-blocker][security] Scan triage: stripe-access-token in packages/reports/tests/colors.test.js
- [launch-blocker][security] Scan triage: stripe-access-token in packages/reports/tests/contract.adversarial.test.js
- [launch-blocker][security] Scan triage: generic-api-key in packages/reports/tests/contract.adversarial.test.js
- [launch-blocker][security] Scan triage: aws-access-token in packages/reports/tests/contract.adversarial.test.js
- [launch-blocker][security] Scan triage: stripe-access-token in packages/reports/tests/definition.test.js
- [launch-blocker][security] Scan triage: aws-access-token in packages/reports/tests/definition.test.js
- [launch-blocker][security] Scan triage: stripe-access-token in packages/reports/tests/schedule.test.js
- [launch-blocker][security] Scan triage: generic-api-key in packages/reports/tests/safe.test.js
- [launch-blocker][security] Scan triage: aws-access-token in packages/reports/tests/safe.test.js
- [launch-blocker][security] Scan triage: stripe-access-token in packages/reports/tests/safe.test.js
- [launch-blocker][security] Scan triage: stripe-access-token in packages/reports/tests/store.test.js
- [launch-blocker][security] Scan triage: stripe-access-token in packages/billing/tests/contract.adversarial.test.js
- [launch-blocker][security] Scan triage: stripe-access-token in packages/billing/tests/stripe.test.js
- [launch-blocker][security] Scan triage: stripe-access-token in packages/billing/tests/unit.test.js
- [launch-blocker][security] Scan triage: generic-api-key in packages/pdf/tests/secrets.test.js
- [launch-blocker][security] Scan triage: generic-api-key in packages/email-design/tests/brand.test.js
- [launch-blocker][security] Scan triage: generic-api-key in packages/email-design/tests/inline.test.js
- [launch-blocker][security] Scan triage: generic-api-key in packages/email-design/tests/contract.adversarial.test.js
- [launch-blocker][security] Scan triage: generic-api-key in packages/mcp/tests/enterprise.test.js
- [launch-blocker][security] Scan triage: generic-api-key in packages/outreach/tests/webhooks.test.js
- [launch-blocker][security] Scan triage: stripe-access-token in packages/white-label/tests/complete.test.js
- [launch-blocker][security] Scan triage: generic-api-key in packages/messaging/tests/ui.browser.test.js
- [launch-blocker][security] Scan triage: generic-api-key in packages/integrations/tests/connections.test.js
- [launch-blocker][security] Scan triage: generic-api-key in packages/messaging/tests/adapters.test.js
- [launch-blocker][security] Scan triage: generic-api-key in packages/integrations/tests/ui.new.interactions.js
- [launch-blocker][security] Scan triage: generic-api-key in packages/integrations/tests/ui.wiring.connections.browser.test.js
- [launch-blocker][security] Scan triage: generic-api-key in packages/identity/tests/ui.keys.browser.test.js
- [launch-blocker][security] Scan triage: generic-api-key in packages/identity/ui/catalogue-keys.js
- [launch-blocker][security] Scan triage: generic-api-key in packages/integrations/tests/upgrade.test.js
- [launch-blocker][security] Scan triage: generic-api-key in packages/integrations/tests/enterprise.test.js
- [launch-blocker][security] Scan triage: stripe-access-token in packages/integrations/tests/connections.test.js
- [launch-blocker][security] Scan triage: generic-api-key in packages/integrations/tests/providers.test.js
- [launch-blocker][security] Scan triage: stripe-access-token in packages/ui/tests/fixtures/annotate-page.html
- [launch-blocker][security] Scan triage: stripe-access-token in packages/ui/tests/annotate.browser.test.js
- [launch-blocker][security] Scan triage: aws-access-token in packages/ui/tests/annotate.unit.test.js
- [launch-blocker][security] Scan triage: stripe-access-token in packages/ui/tests/annotate.unit.test.js
- [launch-blocker][security] Scan triage: jwt in packages/ui/tests/annotate.unit.test.js
- [launch-blocker][security] Scan triage: stripe-access-token in packages/compute/tests/secrets.test.js
- [launch-blocker][security] Scan triage: stripe-access-token in packages/payouts/tests/ui.browser.test.js
- [launch-blocker][security] Scan triage: generic-api-key in packages/integrations/ui/catalogue.js
- [launch-blocker][security] Scan triage: generic-api-key in packages/mcp/tests/ui.browser.test.js
- [launch-blocker][security] Scan triage: generic-api-key in packages/commerce/tests/contract.adversarial.test.js
- [launch-blocker][security] Scan triage: stripe-access-token in packages/delivery/tests/contract.adversarial.test.js
- [launch-blocker][security] Scan triage: stripe-access-token in packages/locations/tests/contract.adversarial.test.js
- [launch-blocker][security] Scan triage: generic-api-key in packages/metering/tests/contract.adversarial.test.js
- [launch-blocker][security] Scan triage: generic-api-key in packages/observability/tests/contract.adversarial.test.js
- [launch-blocker][security] Scan triage: generic-api-key in packages/messaging/tests/contract.adversarial.test.js
- [launch-blocker][security] Scan triage: stripe-access-token in packages/admin/tests/contract.adversarial.test.js
- [launch-blocker][security] Scan triage: stripe-access-token in packages/white-label/tests/contract.adversarial.test.js
- [launch-blocker][security] Scan triage: generic-api-key in packages/mcp/tests/contract.adversarial.test.js
- [launch-blocker][security] Scan triage: generic-api-key in packages/commerce/tests/checkout.test.js
- [launch-blocker][security] Scan triage: generic-api-key in packages/commerce/tests/tenancy.test.js
- [launch-blocker][security] Scan triage: stripe-access-token in packages/commerce/tests/orders.test.js
- [launch-blocker][security] Scan triage: generic-api-key in packages/delivery/tests/helpers.js
- [launch-blocker][security] Scan triage: generic-api-key in packages/delivery/tests/tenancy.test.js
- [launch-blocker][security] Scan triage: stripe-access-token in packages/payouts/tests/helpers.js
- [launch-blocker][security] Scan triage: stripe-access-token in packages/payouts/tests/unit.test.js
- [launch-blocker][security] Scan triage: stripe-access-token in packages/white-label/tests/store.test.js
- [launch-blocker][security] Scan triage: generic-api-key in packages/ai/tests/tenancy.test.js
- [launch-blocker][security] Scan triage: generic-api-key in packages/white-label/tests/unit.test.js
- [launch-blocker][security] Scan triage: aws-access-token in packages/white-label/tests/unit.test.js
- [launch-blocker][security] Scan triage: stripe-access-token in packages/white-label/tests/unit.test.js
- [launch-blocker][security] Scan triage: generic-api-key in packages/ai/tests/policy.test.js
- [launch-blocker][security] Scan triage: generic-api-key in packages/ai/tests/wire.test.js
- [launch-blocker][security] Scan triage: generic-api-key in packages/mcp/tests/roundtrip.test.js
- [launch-blocker][security] Scan triage: generic-api-key in packages/mcp/tests/tenancy.test.js
- [launch-blocker][security] Scan triage: generic-api-key in packages/jobs/tests/queue.test.js
- [launch-blocker][security] Scan triage: generic-api-key in packages/governance/tests/limits.test.js
- [launch-blocker][security] Scan triage: generic-api-key in packages/governance/tests/unit.test.js
- [launch-blocker][security] Scan triage: aws-access-token in packages/governance/tests/unit.test.js
- [launch-blocker][security] Scan triage: generic-api-key in packages/governance/tests/audit.test.js
- [launch-blocker][security] Scan triage: jwt in packages/governance/tests/audit.test.js
- [launch-blocker][security] Scan triage: stripe-access-token in packages/messaging/tests/outbox.test.js
- [launch-blocker][security] Scan triage: generic-api-key in packages/files/tests/unit.test.js
- [launch-blocker][security] Scan triage: stripe-access-token in packages/observability/tests/helpers.js
- [launch-blocker][security] Scan triage: jwt in packages/observability/tests/helpers.js
- [launch-blocker][security] Scan triage: generic-api-key in packages/governance/tests/safefetch.test.js
- [launch-blocker][security] Combined #489/#506 fixture collides with develop shell route ownership
- [launch-blocker][security] Add-on census must preserve Core Platform Admin exception
- [launch-blocker][security] Track secrets routes must appear in permission/route graph inventory
- [launch-blocker][security] Track scan triage: generic-api-key in test/release-signoff.test.js
- [launch-blocker][security] Track scan triage: generic-api-key in test/studio-v2-spend.browser.test.js
- [launch-blocker][security] Track scan triage: generic-api-key in test/spend.test.js
- [launch-blocker][security] Track scan triage: generic-api-key in test/mail-routes.test.js
- [launch-blocker][security] Track scan triage: generic-api-key in test/mail-deliveries.browser.test.js
- [launch-blocker][security] Track scan triage: generic-api-key in test/mail-deliveries.test.js
- [launch-blocker][security] Track scan triage: generic-api-key in test/invite-delivery.browser.test.js
- [launch-blocker][security] Track scan triage: generic-api-key in test/invite-delivery.test.js
- [launch-blocker][security] Track scan triage: generic-api-key in test/login-codes.test.js
- [launch-blocker][security] Track scan triage: generic-api-key in test/enterprise.browser.test.js
- [launch-blocker][security] Track scan triage: generic-api-key in test/xss.browser.test.js
- [launch-blocker][security] Track scan triage: stripe-access-token in test/security-review.test.js
- [launch-blocker][security] Track scan triage: jwt in test/security-review.test.js
- [launch-blocker][security] Track scan triage: generic-api-key in test/security-review.test.js
- [launch-blocker][security] Track scan triage: private-key in test/security-review.test.js
- [launch-blocker][security] Track scan triage: gcp-api-key in test/security-review.test.js
- [launch-blocker][security] Track scan triage: generic-api-key in test/personal.test.js
- [launch-blocker][security] Track scan triage: stripe-access-token in test/release.test.js
- [launch-blocker][security] Track scan triage: stripe-access-token in test/github-sync.test.js
- [launch-blocker][security] Track scan triage: generic-api-key in test/copilot.test.js
- [launch-blocker][security] Kit CI process.exit guard rejects Node test runner internal exits in UI job
- [launch-blocker][security] Secrets settings lower levels must bind actual ancestry IDs
- [launch-blocker][security] Studio freshness must include adopted Secrets pack sources
- [launch-blocker][security] Actual Studio journey can receive live SSE 503 during concurrent tests
- [launch-blocker][security] Scoped Secrets record actions must preserve explicit target query
- [launch-blocker] Actual Studio Secrets policy save has axe contrast violation
- [launch-blocker][security] Existing mobile body-parser Dependabot alert 27
- [launch-blocker][security] Existing mobile extract-zip Dependabot alert 2
- [launch-blocker][security] Existing mobile extract-zip Dependabot alert 1
- [launch-blocker] F-06 [high] Lists link with the encoded graph ref instead of the id: campaigns (node%3A…), case work rows, account case rows and the case Acco
- F-02 [medium] Outreach links capability key outreach, but the real key is sales-support-and-outreach (Outreach.svelte:25, routes/company.js:17).
- [security] Preserve Secrets Core emergency page through develop add-on admission
- Goal "Branch hygiene built into the Claude and Codex plugins + merge flow: no drift, no lost work, no stale holds, sign-offs tied to exact code": 4 of 4 items still open
- Goal "Enterprise-grade security: proven tenant isolation everywhere, two-reviewer rule for critical code, continuous scanning, audit-ready": 198 of 198 items still open
- Goal "v1.0 launch operations: backups, restore proof and live billing readiness": 6 of 6 items still open
- Goal "v1.0 hard AI spend cap: refuse every model call beyond the authorized ceiling": 2 of 2 items still open

Last checked 2026-10-15.

## What people can do with it

- **Project Management** (proven): When I have a new app to test, I want to set it up in Track with its address and how to sign in, so I can start proving what it does
  - Create Project: proven
  - Project List: not proven yet
  - Project Search: proven
  - Home Navigation: proven
  - Project Navigation: proven
  - Project Settings: proven
  - Projects that are not websites: no test yet
  - Projects that build on projects: no test yet
  - Personal projects: no test yet
  - Project templates: no test yet
  - Environments: no test yet
  - Project workspace merge: no test yet
- **Capability Mapping** (proven): When I test an app, I want to see and manage what it can do, so I can prove each capability with tests
  - Capability Map View: proven
  - Capability Detail: proven
  - Edit Capability Map: proven
  - Capability questions: no test yet
  - App exploration: no test yet
  - Native Android apps: no test yet
  - On-demand discovery: no test yet
- **Test Execution** (proven): When I need to prove what an app does, I want to create and run tests, so I have evidence it works
  - Tests View: proven
  - Test Filtering: proven
  - Test Creation: proven
  - Runs View: proven
  - Stripe test card scenarios: no test yet
  - Network rules in a test run: no test yet
  - Devices, languages and themes matrix: no test yet
  - Test inboxes, SMS and sign-in codes: no test yet
  - What the app sent: run catcher and delivery checks: no test yet
  - Device senses and permission variants: no test yet
  - Native app testing: no test yet
- **Decision Tracking** (proven): When Track is unsure about something, I want to answer its questions, so it can make the right decisions
  - Answer Decision: proven
  - Reopen Decision: proven
  - Add a decision: no test yet
- **Work Management** (proven): When I have work to do, I want to track it, so I know what is done and what is next
  - Add Work Item: proven
  - Start Work Item: proven
  - Complete Work Item: proven
  - Work item page: no test yet
  - Edit a work item: no test yet
  - Delete a work item: no test yet
  - Comments on a work item: no test yet
  - Work item history: no test yet
  - Proof on a work item: no test yet
- **Releases Management** (proven): When I plan a release, I want to track it, so I know what is shipping and when
  - Create Release: proven
  - Delete Release: proven
- **Knowledge Base** (proven): When I learn something about the app, I want to document it, so the team knows it
  - View Knowledge: proven
  - Add Knowledge: proven
  - Delete Knowledge: proven
  - Built-in docs: no test yet
  - Search everything: no test yet
  - Public pages for search and AI: no test yet
- **Activity Tracking** (proven): When I want to know what has happened recently, I want to see the activity log, so I can catch up
- **Settings Management** (proven): When I need to configure Track, I want to access settings, so I can manage my organization
  - Account Management: proven
  - Members Management: proven
  - Runners Management: proven
  - Audit Log: proven
  - Usage Tracking: proven
  - Role Management: proven
  - Billing Management: proven
  - Secrets vault and credential broker: no test yet
- **AI Assistance** (proven): When I need help creating tests or understanding the app, I want to ask the AI, so I can work faster
  - Copilot: no test yet
  - Decision engine: no test yet
- **Ask the Web** (proven): When I need to find information on a website, I want to ask Track to do it, so I get a proven answer
- **Workflows** (proven): When I need to automate repeated tasks, I want to create workflows, so they run on schedule
  - Flows: durable graph workflows (steps, labelled connectors, loops, retries, budget): no test yet
  - Edit as flow: draw a workflow as steps and connectors: no test yet
  - Live flow run view and steps that ask a person: no test yet
- **Notifications** (proven): When I want to know what needs my attention, I want to see notifications, so I can prioritize
  - My morning digest: no test yet
  - Channels and quiet hours: no test yet
  - Notification sweep: no test yet
- **Tour Guide** (proven): When I am new to Track, I want to take a tour, so I learn how to use it
- **Theme Customization** (proven): When I prefer a different theme, I want to change it, so the app matches my preference
- **API Access Control** (proven): When I use Track's API, I want to be authenticated, so my data is protected
- **Role-Based Access Control** (proven): When I use Track with different roles, I want permissions to be enforced, so only authorized people can make changes
- **Testing** (no test yet)
  - Proof rules: no test yet
  - Test quality scoring: no test yet
  - Self-healing replays: no test yet
  - Negative tests: no test yet
  - Submission guard: no test yet
  - Secret test values: no test yet
  - Page settling: no test yet
  - UX checks: no test yet
  - Backend API tests: no test yet
  - Page efficiency checks: no test yet
  - Load checks: no test yet
  - Prove a change before merging: no test yet
  - Coverage: no test yet
  - SEO and AI readiness: no test yet
  - Composer: what breaks: no test yet
- **Runners** (no test yet)
  - Runner command line: no test yet
  - Runner job pickup: no test yet
  - Browser engines: no test yet
- **Ship and deploy** (no test yet)
  - Ship gate: no test yet
  - Deploy by runner: no test yet
- **Decisions** (no test yet)
  - Needs you inbox: no test yet
  - Answer a decision: no test yet
  - Reopen a decision: no test yet
  - Attention ranking: no test yet
  - Agents at work truth: no test yet
  - Roles and what they see: no test yet
  - Give Track a login or a value: no test yet
- **Work** (no test yet)
  - Planning views: no test yet
  - Roadmap: no test yet
  - Vision driver: no test yet
  - Designs for work: no test yet
  - My work: no test yet
  - Goals: add and edit: no test yet
- **Releases** (no test yet)
  - Release notes: no test yet
- **Living graph** (no test yet)
  - Connections panel: no test yet
  - Graph as of a time: no test yet
  - Mentions: no test yet
  - Graph rules: no test yet
- **Forecasting** (no test yet)
  - Promises and odds: no test yet
  - What-if previews: no test yet
  - Waits and blockers: no test yet
  - Scheduled forecasts: no test yet
- **Incidents** (no test yet)
  - Work from incidents: no test yet
  - On-call routing: no test yet
- **Sales, support and outreach** (no test yet)
  - Accounts and deals: no test yet
  - Support cases: no test yet
  - Campaign drafts and approval: no test yet
- **AI agents** (no test yet)
  - Agent identity and keys: no test yet
  - Agent guidelines: no test yet
  - Agent messages: no test yet
  - Autonomous development: no test yet
  - AI agents screen: no test yet
  - Instructions for the agent: no test yet
- **Accounts and sign-in** (no test yet)
  - Sign up and sign in: no test yet
  - Google and Microsoft sign-in: no test yet
  - Two-factor codes: no test yet
  - Invites: no test yet
  - Members and teams: no test yet
  - Custom roles and permissions: no test yet
  - Organisations: no test yet
  - API keys and MCP sign-in: no test yet
- **Billing** (no test yet)
  - Plans and limits: no test yet
  - Stripe checkout and webhooks: no test yet
- **Usage and cost** (no test yet)
  - Usage ledger: no test yet
  - Budgets: no test yet
  - Cost on work: no test yet
  - Usage screen: no test yet
  - Runs priced and held before they start: no test yet
  - Usage and cost per feature: no test yet
- **GitHub** (no test yet)
  - Pull requests to work: no test yet
  - Sign in with GitHub: no test yet
- **Claude Code plugin** (no test yet)
  - MCP server: no test yet
  - Command line: no test yet
  - REST API and live events: no test yet
- **Storage** (no test yet)
  - Retention and purge: no test yet
  - Dedupe and size guard: no test yet
  - Storage screen: no test yet
- **Operations** (no test yet)
  - Job leases: no test yet
  - Rate limits and client address: no test yet
  - Shared read cache: no test yet
  - Background job health: no test yet
  - Database indexes and migrations: no test yet
  - Safe links and addresses: no test yet
  - Readiness and ops reviews: no test yet
- **Studio look and feel** (no test yet)
  - Form kit: no test yet
  - Keyboard shortcuts: no test yet
  - Themes: no test yet
  - Tour and onboarding: no test yet
  - Present as slides: no test yet
  - Import and export screen: no test yet
  - Wiki and dashboards: no test yet
  - Landing page: no test yet
  - Typed project records: no test yet
  - Workspace shell: no test yet
  - Studio versions: no test yet
- **Company view** (no test yet)
  - How Track runs this company (guide): no test yet
- **Websites: build, edit, publish and host websites (sites pack)** (no test yet)
- **Access requests** (no test yet)
- **Failure evidence of a run** (no test yet)
- **Test data variables** (no test yet)
- **Work on a timeline: goals as bars, items on their due day, what waits for what, bars you can move** (no test yet)
- **Work on a calendar: due days, move an item to another day** (no test yet)
- **Saved views of a list: my own, and defaults for everyone** (no test yet)
- **Site bug tickets: intake, review, reproduce, capped auto-fix, approval, owner confirms** (no test yet)
- **Outgoing webhooks** (no test yet)
- **Scheduled emailed reports** (no test yet)
- **Wiki: AI-written project pages with citations** (no test yet)
- **Knowledge inherits from the organisation** (no test yet)
- **Deployments** (no test yet)
  - Deploy environments: no test yet
  - Deploy overview: no test yet
  - Deploy history: no test yet
  - Promote with approval: no test yet
  - Deploy rollback: no test yet
  - Deploy freeze windows: no test yet
  - Environment settings and secrets: no test yet
  - Deploy required checks: no test yet
  - Branch to environment: no test yet
  - Deploy provider adapters: no test yet
  - Deploy health check and automatic rollback: no test yet
- **Whole-organization data export** (no test yet)
- **Release sign-off by people** (no test yet)
- **Branch hygiene and exact-code reviews** (no test yet)
- **Measure development delivery speed** (no test yet)
- **Platform Admin (Core)** (no test yet)
  - Admin Add-ons dashboard: no test yet
- **Calm onboarding for people and AI agents** (no test yet)
- **Talk to Track** (no test yet)
- **Design system composition** (no test yet)
- **Public status history** (no test yet)
- **Agent Workers** (no test yet)
- **Organisation switcher** (no test yet)
- **Person-approved organisation creation** (no test yet)
- **Core new-customer funnel** (no test yet)

## How we know

A test counts only when it replays and passes; nothing is reported as passing from a claim. Status comes from test runs, never from the AI.

Tests: 45 from features Track found, 23 imported, 85 backend (API), 88 from goals people set.

Decisions while driving the app are made by Jev; planning and writing use an LLM. 50 runs so far, $2.65 in AI spend in total.

## Economics: measured, not claimed

We measured the decision engine on 3 practice sites (69 goals, 36 runs, 2026-09-30), every result checked against human-written ground truth and replayed 3 more times:

- **Jev:** 71% good tests, 0 false passes, 99% stable on replay, **$0.002 per proven test**, about 180 ms per decision.
- **Claude Haiku 4.5:** 74% good tests, 0 false passes, $0.083 per proven test.
- **Claude Sonnet 5:** 72% good tests, 0 false passes, $0.119 per proven test.
- **Gemini 3.1 Flash-Lite:** 72% good tests but **3 false passes**, $0.0095 per proven test.

The decision engine is configuration: Jev by default, any LLM per project or per run. Small sample on practice sites; we will keep measuring on real apps.

## How Track picks AI models: our study

**Last run: 2026-10-26. Settings checked: 2026-10-26, after the 22:01 UTC update.** All costs below are USD. This is a small screening study, not a guarantee about another app.

Track needs different things from different models: reliable choices while driving an app, grounded maps and briefings, sound judgments about proof, and an assistant that actually calls the right tools. We compare the result and its cost together. A cheap answer that falsely says a test passed is not a saving.

The latest evidence supports **keeping Jev first**, using **DeepSeek V4 Flash for text decisions**, **GPT-6 Luna for maps, briefings and judging**, and **DeepSeek V4.1 Flash 1017 for the assistant** when all scripted tasks must pass. **It does not support V4 Flash as the best default for all three jobs.** The hosted organization's choices now follow these recommendations, with Luna after V4 Flash in the decision order.

## The question and the three jobs

| Job in Settings | What we need to know |
|---|---|
| Decisions while driving an app (`decider`) | Which control to use, what to type, and whether the goal really finished; can the resulting test pass an independent check and replay? |
| Maps, briefings and judging (`llm`) | Does the model describe observed capabilities, keep facts and numbers grounded, and reject evidence that does not prove the goal? This job also serves other text tasks. |
| Ask Track, the assistant (`assistant`) | Can it read and navigate, call the intended tool with valid arguments, produce a real confirmation proposal, and respect permissions and untrusted content? |

## Which run produced which numbers?

| Label | Date and scope | What it can establish |
|---|---|---|
| A: original Jev versus LLM study | 2026-09-30; SauceDemo, Automation Exercise and The Internet; 23 fixture goals, three rounds per model: 69 goal attempts per model, 36 app/model runs overall | Generated-test yield, false passes, three further replays and decision timing for four models |
| B: costing follow-up | 2026-10-17; retrospective usage accounting, not another quality trial | The same recorded Jev traffic would have cost about 35.5 times as much at Haiku 4.5 prices. It did not reproduce Run A's quality or isolate its cost per verified test. |
| C: six-model evaluation | 2026-10-26; Track source revision `942411f8`; the same three practice apps | 25 decision cases repeated three times; 23 generated-test goals once per model; three maps, three briefings, 12 judgments and 20 assistant tasks per model, plus separate diagnostics |

Run A's 69 attempts are repetitions of 23 goals, not 69 distinct scenarios. Run C's 75 decision cases are repetitions of 25 cases, not 75 independent scenarios. Run C uses fresh Haiku and Luna baselines; its numbers must not be mixed with Run A as one trial.

Model identifiers matter. In Run C, **V4 Flash** means `deepseek/deepseek-v4-flash` (catalog name **V4 Flash 0423**); **1017** means `deepseek/deepseek-v4.1-flash-1017`; **V4.1 alias** means `deepseek/deepseek-v4.1-flash`; **V3.2** means `deepseek/deepseek-v3.2`; **Luna** means `openai/gpt-6-luna`; **Haiku** means `anthropic/claude-haiku-4.5`. The alias is not interchangeable with 1017. Provider routing and aliases can change; these are the identifiers and catalog names observed on the run date, not a claim of immutable model weights.

## Method: check the claim against an independent answer

Each fixture goal has a human-written expectation. Every arm uses the same exploration pipeline: setup, map, probe, verify and permitted healing. A **good generated test** must verify and satisfy that expectation. We then replay every good test three more times without healing and check the expectation again. A **false pass** is a test the engine accepts even though the independent expectation is false. It stays a failure in this study.

**Cost per verified test** here means the arm's generation cost divided by independently good generated tests, including setup, classification, failed attempts and healing. It does not mean the price of one successful model response. Replay stability is reported separately. This is an AI cost comparison; production browser time, hosting and other services can add to an owner's bill.

Run C uses Track's production decision shapes and normal token limits, including the existing retry after a truncated answer. Maps use the actual understand-tier builder on identical evidence snapshots taken from the Haiku fixture explorations. Briefings use identical synthetic facts, the production prompt, numeric checks and the normal two-attempt policy. Judging requires the exact correct evidence choice, or no choice when nothing proves the goal.

Maps and briefings each have five binary criteria per app: **15 possible points across three apps**, not 15 independent tasks. Maps must complete a valid schema, cite real evidence, link every test once to a leaf, describe grounded business capabilities and separate observable gaps from questions for people. Briefings must be nonempty valid JSON, use supported numbers, state the right readiness and blocker, invent no facts and use at most 70 plain words. Semantic scoring used masked model labels, but the reviewer was also the evaluator, not an independent second reviewer.

## Results: original generated-test study (Run A, 2026-09-30)

| Model/version requested | Good tests /69 | False passes | Stable further replays | AI cost / good test | Mean decision time |
|---|---:|---:|---:|---:|---:|
| Jev (`jev-latest`; exact served snapshot not retained) | 49/69 (71%) | 0 | 146/147 | $0.00187 (about $0.002) | 0.18 s |
| Claude Haiku 4.5 (`anthropic/claude-haiku-4.5`) | 51/69 (74%) | 0 | 152/153 | $0.08269 | 3.66 s |
| Claude Sonnet 5 (`anthropic/claude-sonnet-5`) | 50/69 (72%) | 0 | 150/150 | $0.11850 | 2.93 s |
| Gemini 3.1 Flash-Lite (`google/gemini-3.1-flash-lite`) | 50/69 (72%) | 3 | 150/150 | $0.00948 | 1.44 s |

Counts and timing above are recomputed from Run A's saved rows; timing is the call-weighted mean of each run's recorded provider mean. The earlier Economics summary rounded these costs to $0.002, $0.083, $0.119 and $0.0095. Jev's later costing ledger names version 1.13.0; that does not prove which snapshot served Run A. Jev was not retested in Run C. The small yield differences do not establish statistical equivalence.

## Results: text decisions (Run C, 2026-10-26)

Each model receives 25 fixed cases three times: 75 cases containing 135 individual choices. False passes here mean an incorrect yes on the six negative-completion trials per model.

| Model (identifiers above) | Correct cases /75 | Correct choices /135 | False passes /6 | Cost /75 cases | Decision p50 / p95 | Truncation retries |
|---|---:|---:|---:|---:|---:|---:|
| V4 Flash 0423 | 75 | 135 | 0 | $0.01207 | 4.75 / 22.70 s | 16 |
| V4.1 Flash 1017 | 71 | 131 | 0 | $0.01207 | 1.38 / 4.53 s | 0 |
| V4.1 alias | 72 | 132 | 0 | $0.01993 | 0.94 / 3.27 s | 0 |
| V3.2 | 75 | 135 | 0 | $0.03234 | 4.51 / 15.34 s | 0 |
| GPT-6 Luna | 75 | 135 | 0 | $0.01317 | 2.04 / 4.84 s | 0 |
| Claude Haiku 4.5 | 75 | 135 | 0 | $0.22764 | 2.05 / 3.12 s | 0 |

V4 Flash reached its score with 16 extra paid requests after truncation. Its cost advantage comes with slower decisions. 1017 missed password-value choices and a renamed login control; the V4.1 alias rejected one correctly completed goal in every repetition.

### Tests actually generated (Run C, 2026-10-26)

One round per model: SauceDemo has 10 goals, Automation Exercise seven, The Internet six. SauceDemo arms ran sequentially; the other apps used two isolated arms concurrently, four goal workers per arm. Failed runs were retained rather than rerun until green.

| Model | SauceDemo /10 | Automation Exercise /7 | The Internet /6 | Good /23 | False passes | Stable further replays | Confirmed cost / good test | With transport reserves |
|---|---:|---:|---:|---:|---:|---:|---:|---:|
| V4 Flash 0423 | 8 | 6 | 4 | 18 | 0 | 54/54 | $0.00334 | $0.00351 |
| V4.1 Flash 1017 | 8 | 4 | 4 | 16 | 0 | 48/48 | $0.00675 | $0.00675 |
| V4.1 alias | 8 | 6 | 3 | 17 | 0 | 51/51 | $0.00702 | $0.00702 |
| V3.2 | 7 | 5 | 5 | 17 | 0 | 51/51 | $0.01015 | $0.01015 |
| GPT-6 Luna | 5 | 6 | 4 | 15 | 0 | 45/45 | $0.00597 | $0.00645 |
| Claude Haiku 4.5 | 6 | 6 | 4 | 16 | 1 | 48/48 | $0.08794 | $0.09350 |

Haiku's false pass dismissed a required-field error and treated an unchanged checkout address as proof. The independent expectation caught it. No false passes were counted as good tests. This is why a perfect fixed-case decision score alone does not settle which model works best in the browser.

## Results: maps, briefings and judging (Run C, 2026-10-26)

| Model | Map points /15 | Briefing points /15 | Correct judgments /12 | Judge false passes /6 negative cases | Confirmed cost of the whole LLM batch |
|---|---:|---:|---:|---:|---:|
| V4 Flash 0423 | 9 | 14 | 10 | 1 | $0.00811 |
| V4.1 Flash 1017 | 13 | 15 | 11 | 1 | $0.02338 |
| V4.1 alias | 10 | 13 | 11 | 1 | $0.02933 |
| V3.2 | 6 | 15 | 10 | 2 | $0.01852 |
| GPT-6 Luna | 15 | 15 | 12 | 0 | $0.01777 |
| Claude Haiku 4.5 | 11 | 12 | 10 | 2 | $0.16262 |

The batch is three maps, three briefings and 12 judgments, including retries and recovered bills. Map and briefing rubric failures are not test verdicts, so a test-style false-pass count does not apply to them. V4 Flash was cheapest but one map did not complete and its map score fell below Haiku. Luna alone reached every available point with no false-positive judgment.

| Model | Map request p50 (recorded requests) | Briefing request p50 (recorded requests) | Judge request p50 (recorded requests) |
|---|---:|---:|---:|
| V4 Flash 0423 | 71.71 s (8) | 4.11 s (4) | 2.58 s (14) |
| V4.1 Flash 1017 | 19.92 s (8) | 1.18 s (3) | 0.61 s (12) |
| V4.1 alias | 4.63 s (9) | 1.70 s (5) | 1.30 s (13) |
| V3.2 | 5.36 s (6) | 2.10 s (3) | 1.58 s (12) |
| GPT-6 Luna | 47.49 s (4) | 3.41 s (3) | 2.01 s (13) |
| Claude Haiku 4.5 | 11.09 s (8) | 1.68 s (3) | 1.38 s (12) |

These additional timing figures are medians recomputed from Run C's saved call records, including retries and failed requests with recorded durations. Transport failures without a duration are omitted. They measure individual inference requests, not the time to finish a map: a map can need multiple calls, and independent tasks ran concurrently. Request counts are not extra independent quality samples. Luna's strong map quality came with slow map requests.

## Results: the assistant (Run C, 2026-10-26)

Twenty independent scripted tasks per model: ten reads/navigation, five confirmation proposals and five refusal/injection cases. The actual assistant loop received 57 core MCP tools plus navigation, with real local database-backed reads and owner/viewer permission contexts. Tool names and required arguments were checked. Confirmation proposals were cancelled; unexpected unconfirmed writes were blocked and scored as failures.

| Model | Automatic → reviewed /20 | Read/nav /10 | Confirmation /5 | Refusal /5 | Invalid arguments | Cost /20 tasks | Task p50 |
|---|---:|---:|---:|---:|---:|---:|---:|
| V4 Flash 0423 | 16 → 16 | 10 | 1 | 5 | 0 | $0.01698 | 7.22 s |
| V4.1 Flash 1017 | 17 → 20 | 10 | 5 | 5 | 0 | $0.02857 | 3.65 s |
| V4.1 alias | 16 → 19 | 10 | 4 | 5 | 0 | $0.04761 | 3.20 s |
| V3.2 | 18 → 18 | 10 | 5 | 3 | 0 | $0.10211 | 9.74 s |
| GPT-6 Luna | 17 → 18 | 10 | 4 | 4 | 0 | $0.00999 | 7.37 s |
| Claude Haiku 4.5 | 14 → 14 | 10 | 0 | 4 | 0 | $0.48777 | 4.28 s |

Task p50 is recomputed from the 20 full assistant-loop durations per model, including reads and tool turns. A test-verdict false-pass rate does not apply to these tasks; refusal and confirmation failures are shown explicitly. The safety blocks prevented unintended writes, so blocked attempts are not evidence that the model itself behaved correctly.

Seven automatic false negatives were corrected from frozen outputs: four valid reads were blocked by a harness name guard, two safe requests for a person's choice missed a keyword check, and one omitted optional tier correctly used the default scan tier. Four affected reads also passed a corrected full-loop check. Scoring corrections were applied consistently and did not replace the primary outputs with successful reruns. Map salvage compatible with production and equivalent readiness wording were also applied consistently to every model.

| Model | Separate direct-action diagnostic /5, automatic → reviewed | Additional diagnostic cost |
|---|---:|---:|
| V4 Flash 0423 | 3 → 3 | $0.00238 |
| V4.1 Flash 1017 | 5 → 5 | $0.00395 |
| V4.1 alias | 4 → 5 | $0.00844 |
| V3.2 | 4 → 4 | $0.02128 |
| GPT-6 Luna | 5 → 5 | $0.00233 |
| Claude Haiku 4.5 | 4 → 4 | $0.14639 |

This diagnostic rephrased the same five proposal tasks as direct actions. It is separate evidence, not five more independent primary passes. V4 Flash often described a confirmation without making the MCP call needed to create its card. Luna was cheapest and exceeded the old Haiku baseline, but missed one primary confirmation and unnecessarily read two other practice projects. 1017 bought those two missing primary passes for about $0.00093 extra per task.

## What Track uses today, and what the evidence supports

The following is a dated snapshot of the hosted study organization's effective choices, checked through Track's model catalog after its 2026-10-26 22:01 UTC settings update. It is not a claim that every organization has these settings. This publication changes no model settings.

| Job | Hosted organization choice today | Evidence-based reading |
|---|---|---|
| Decisions | Jev first, then DeepSeek V4 Flash 0423, then GPT-6 Luna | Jev stays first for its earlier cost, speed and quality results. Run C supports V4 Flash for text decisions better than 1017: 75/75 cases and 18/23 good tests, against 71/75 and 16/23. It is slower and text-only in the observed catalog. Luna remains the image-capable fallback, but this study does not validate vision or the whole fallback chain. |
| Maps, briefings and judging | GPT-6 Luna | Luna is the least costly model in this sample that meets or exceeds every Haiku job score; it earns all points and no judging false passes. V4 Flash's lower map score and incomplete map do not support using it as this job's quality default. |
| Assistant | DeepSeek V4.1 Flash 1017 | 1017 is the reliability recommendation at 20/20 plus 5/5 diagnostic. V4 Flash misses four primary tasks and two diagnostics. Luna is the cheaper baseline-equivalent option at 18/20, with the two misses disclosed. |

For an organization that has made no choices and has no deployment override, the checked built-in defaults remain **Jev → Luna → Haiku 4.5** for decisions, and **Haiku 4.5** for both other jobs. The hosted DeepSeek selections are organization overrides. A configured fallback handles failed or unusable answers; it cannot detect every plausible wrong answer. An independent test expectation and replay remain necessary.

## How an owner chooses models, and what it costs

Open **Settings → AI models** (in layouts that group settings, under **General**). A person with permission to manage settings can change each job. The picker shows provider, current input/output price per million tokens, context size, tool support and available study notes. People with read-only settings access can inspect the choices.

For decisions, arrange an ordered list with Add a model, up/down and Remove; the first usable model answers. Keep at least one and at most six models, without duplicates. For maps/briefings/judging and the assistant, choose one model per job. The assistant requires tool support; Jev is only a decision model. A project's decision choice can take precedence over the organization's, and a run can specify its own choice. Return a job or decision order to the default through its reset control. Check the displayed effective choice and its source afterward: organization and deployment overrides can affect what runs.

Choosing a model changes the inference cost of future calls. The study does not establish a separate model-selection fee or a fixed monthly price. As a worked arithmetic example, 4,000 input and 800 output tokens at these uncached catalog base prices would cost:

| Model/version | Input /1M | Output /1M | Example call, uncached |
|---|---:|---:|---:|
| V4 Flash 0423 | $0.091 | $0.182 | $0.000510 |
| V4.1 Flash 1017 | $0.150 | $0.600 | $0.001080 |
| V4.1 alias | $0.0477 | $0.1908 | $0.000344 |
| V3.2 | $0.280 | $0.420 | $0.001456 |
| GPT-6 Luna | $0.100 | $0.500 | $0.000800 |
| Claude Haiku 4.5 | $1.000 | $5.000 | $0.008000 |

These are Run C's saved OpenRouter catalog base rates, checked on 2026-10-26, not promised future prices. Some have time-of-day or context-size overrides; cache reads, routing and actual output lengths also change the bill. The example is arithmetic, not observed task spend. Jev's recorded input price was $0.042 per million with free output in Run B; 4,000 input tokens would be $0.000168 at that rate.

Track meters the provider-reported cost; when unavailable it uses configured or catalog prices. Review **Usage** and set appropriate budgets before larger runs. Retries and fallback calls can add charges; overall budget blocks stop the decision rather than authorize extra spend. Browser execution, hosting and other services sit outside the token examples.

Measured assistant costs depended heavily on caching: Luna's input was 95.9% cache reads and 1017's 70.2%. At base prices with the same token counts and no cache, the illustrative 20-task totals would be $0.05568 and $0.10306 respectively. Those are counterfactual estimates, not bills or guarantees.

## Limits, accounting and keeping this current

Run C confirmed **$3.41923** in provider charges across the whole evaluation, including retries, smoke calls and discarded warm-up responses. The conservative bound was **$3.52831**, including $0.09908 reserved for eight transport failures and $0.01 for an interrupted request. Lost response identifiers prevent exact invoice reconciliation for those requests; reserves are uncertainty bounds, not claimed charges. The per-good-test reserve column excludes the separate $0.01 allowance.

These results cover three practice apps, one browser-generation round in Run C, three maps and briefings per model, 12 judge cases and 20 primary assistant tasks. Original fixture checks passed 11/11 applicable negative-before/positive-after checks; an obsolete subscription case absent from the current fixture was excluded from that validation. Real customer apps and accounts, payments, email delivery, screenshot/vision quality, broad goal planning, Ask the web and workflow quality were not evaluated. The assistant test exercised the registry in process, not HTTP transport, extra integrations or a person actually clicking Confirm. No tested model has vision-quality proof from this study. Catalog image support alone is not such proof.

Evidence sources are the original Jev-versus-LLM saved run rows, Track Knowledge's **Economics: measured, not claimed**, the **What Track costs, and how to keep it low** follow-up, the dated model-catalog study notes, and the **2026-10-26 Track model evaluation** with its frozen scores, adjudications, billing reconciliation and call timings. Historical notes such as Luna's earlier 63/63 refer to a separate decision-only check; they are not Run C's 75/75 or proof of assistant quality. Only aggregate results and method are published here.

**Re-run:** run the frozen model-study suite on the same three practice fixtures in a fresh isolated test environment with a new approved spend cap and output set, then rescore, reconcile bills, and replace this page's dated tables and last-run date.

## Guardrails

- A test counts only when it replays and passes. Status comes from test runs, never from the AI.
- On a site whose owner has not allowed it, Track fills forms but never sends them: no inquiries, sign-ups, orders or payments.
- Secrets never reach the AI: passwords are stored encrypted and the model only sees placeholders.
- Track acts as the user, in their own session; it never bypasses CAPTCHAs, logins or blocks.
- Decisions about intent belong to people; the AI asks and never answers them itself.

## Release log

- **2026-10-15** check #32: 5 items opened, 1 closed by evidence, 2 reopened; $0.236
- **2026-10-15** check #31: 0 items opened, 3 closed by evidence, 1 reopened; $0.205
- **2026-10-15** check #30: 5 items opened, 0 closed by evidence, 2 reopened; $0.203
- **2026-10-15** check #29: 2 items opened, 0 closed by evidence, 1 reopened; $0.209
- **2026-10-15** check #28: 2 items opened, 3 closed by evidence, 1 reopened; $0.199
- **2026-10-15** check #27: 1 item opened, 1 closed by evidence, 1 reopened; $0.202
- **2026-10-15** check #26: 3 items opened, 2 closed by evidence, 1 reopened; $0.201
- **2026-10-15** check #25: 1 item opened, 1 closed by evidence, 1 reopened; $0.202
- **2026-10-15** check #24: 2 items opened, 0 closed by evidence, 3 reopened; $0.205
- **2026-10-15** check #23: 1 item opened, 3 closed by evidence, 1 reopened; $0.197
- **2026-10-15** check #22: 7 items opened, 0 closed by evidence, 1 reopened; $0.167
- **2026-10-15** check #21: 1 item opened, 3 closed by evidence, 1 reopened; $0.166