Clones dashboard

The Molar Clones dashboard is your control surface for spawn lifecycle, session inspection, chaos presets, and fleet health. Access it at:

app.molar.it/dashboard/clones

Product marketing and positioning: clones.molar.it

The dashboard is available in standalone mode (Clones-only SKU) and combined mode (full Molar bundle with shared org session from app.molar.it).

Information architecture

Naming: Clones uses Sessions in the UI (not "Runs") to avoid collision with Guard Checks and Cartographer Agent runs.

Fleet overview (Home)

Single pane for clone capacity:

• Active runs count by kind (stripe / twilio / email / auth / s3)
• Concurrent run quota vs plan limit
• Recent spawns/teardowns with actor (user, CI, MCP agent, Guard PR)
• Health rollup per clone kind
• Destructive-call refusal counter — safety KPI showing blocked live-vendor attempts
• Quick actions: Spawn new run, resume last world snapshot, copy CLI snippet
• On-prem banner (when applicable): Helm version, Garage snapshot store status

Sessions list

Entry point for all clone runs.

Filters: clone kind(s), status (active / paused / expired / teardown), source (local SDK, hosted spawn, Guard PR, MCP, CLI)

Columns: runId, kinds active, seed, clock offset, request count, snapshot count, TTL remaining

Bulk actions: teardown selected, extend TTL, export snapshot bundle

Linkages:

• Guard PR runs → link to Guard check
• Molar scenarios → scenario id/slug

Search: runId, seed, metadata tags

Run workspace

The primary daily-use surface for one isolated test world (/sessions/{runId}).

Connection panel

• API base URL, admin base URL, per-kind URLs
• Copy SDK env block (MOLAR_CLONE_STRIPE_URL, etc.)
• Per-run JWT mint/regenerate (1h, admin scope)
• Seed, persistence mode, schema version
• Clock state: frozen at, offset, paused indicator
• Spawn source, expiry countdown, extend/teardown

Live request stream

Tail NDJSON request logs (SSE). Filter by clone kind, path, status, error_injected. Click a row for full request/response. Export as JSON Lines for CI artifacts.

Webhook inspector

Configured endpoints, delivery attempts, signature headers (Stripe-Signature, X-Twilio-Signature, svix-*), retry schedule, force redelivery, golden signature comparison.

Virtual clock control

Advance by duration (30d, 2h), set absolute time, pause/resume. Scheduled events preview shows what fires on next advance (subscription renewal, invoice finalize, OTP expiry). Sync world clock advances all registered clones together.

Chaos / fault injection

Active rules list, create rule (path, method, status, latency, rate). Presets library:

Snapshots

List, save (with label), restore, download gzip, delete. Compare two snapshots for resource diff. Branch from snapshot — spawn new runId initialized from snapshot.

World snapshot

Register/unregister clone admin URLs. World snapshot → manifest { stripe: ref, twilio: ref, ... }. World restore with per-component progress.

Destructive-call safety panel

Live feed of refusal events, misconfiguration hints (env still pointing at api.stripe.com), strict mode toggle, test refusal button.

Integration shortcuts

• Open in Cartographer Trace
• Open Guard failure (if PR-linked)
• Copy Playwright fixture / MCP invocation snippet

Vendor explorers

Postman-like state browsers at /clones/{kind}:

• Run selector (defaults to last active)
• Resource tree/table by vendor object type
• Cross-clone links (Stripe invoice PDF → S3 object; Auth reset → Email inbox)
• Endpoint coverage indicator (% ofAPI coverage)
• 501 Not Implemented log

Clone preset library (/library)

One-click onboarding — don't start from empty state:

Save org-private presets (snapshot + chaos rules). Import/export JSON bundles for air-gap.

Fleet & BYOC (/fleet)

See BYOC & fleet for deployment details. Dashboard shows:

• Registered BYOC endpoints (kind, API URL, admin URL, health)
• Helm release status (on-prem)
• Pre-warmed pool config (cloud)
• Air-gap audit timestamp
• Garage / snapshot store health

Observability (/observability)

• Prometheus metrics: requests/sec, p99 latency, error rate, active tenants
• OTLP trace links to Cartographer/Tempo
• Per-kind SLO: cold-start p99 ≤ 2s, spawn success rate
• Tenant isolation audit results
• Webhook golden-test CI status

Combined-mode chrome

When your org has the full Molar bundle:

• Org switcher and Molar home → app.molar.it
• App switcher (Guard | Cartographer | Clones | Trace)
• Team/SSO from central org identity
• Product usage on Clones billing; Org billing links to central view

What the dashboard does not do

• Guard PR gating UI (links to Guard instead)
• Cartographer trace debugger (links to Trace)
• Scenario compile/validate (links to app.molar.it /scenarios)
• Generic HTTP stub editor
• Full Kubernetes cluster dashboard

What's available today

Fleet & BYOC: connect self-hosted clone endpoints from Settings → Fleet. See BYOC & fleet.

If a control is grayed out, your plan may not include it yet — check app.molar.it/billing or contact support@molar.it.