Local-first MQTT workbench

Debug MQTT from live trace to replayable proof.

Mqttable turns broker tabs, MQTT clients, TraceGrid rows, PCAP diagnostics, Replay scenarios, Proxy faults, Bench runs, Monitor evidence, CLI, and MCP into one local workflow.

Free local runtime first. Pro only expands scale, depth, Remote Runtime, and advanced agent workflows.

Local

Desktop runtime and data stay on your machine by default.

Trace

Live rows, PCAP findings, Replay reports, and Monitor events stay linked.

CLI/MCP

Agents read schemas, prepare plans, apply only after confirmation, then verify.

Live MQTT workspace

The app surface mirrors the real Mqttable workflow.

Local Runtime
Local broker301QUIC target98TLS staging42

Clients

3 online

Trace rows

1,284

Replay ready

12 steps

Operate clients, keep trace context, replay from evidence, and hand the same state to CLI or MCP when automation helps.

Time
Client ID
Type
Direction
Topic
QoS
Payload / reason
11:30:21
mqttable-e25b
PUBLISH
← receive
test/1/2
0
undefined
11:29:51
mqttable-e25b
PINGRESP
← receive
none
-
keepalive confirmed
11:28:44
replay-run-184
PUBACK
→ verify
replay/verification
1
expected response matched
11:27:10
bench-group-a
METRIC
↔ trend
bench/preflight
-
host capacity evidence attached
mqttable-e25b4401topic test/1/2TEXT · Empty payloadSend

Why the old tools break

MQTT incidents lose truth when every tool owns one fragment.

A real issue crosses live clients, packet traces, capture files, timing, network faults, load, local runtime pressure, and a handoff to another engineer or agent. Mqttable keeps that chain visible.

01

Connection context disappears

A client reconnects, a retained message changes, a subscription moves, and the screenshot no longer explains what happened.

02

Captures do not explain themselves

Raw PCAP rows are useful, but the diagnostic checks, issue groups, traffic summary, and source references need to stay attached.

03

Resend is not reproduction

You need a named Replay Scenario with mapped identities, topic rewrites, pace, expected protocol responses, and a report.

04

Numbers without host evidence mislead

Bench results need preflight, runtime metrics, history, comparison, and report evidence next to the chart.

What is actually in Mqttable

One workbench for the boring, crucial MQTT work.

The landing page should sell the product that exists in ../mqttable: local desktop, Phoenix LiveView surface, TraceGrid, PCAP diagnostics, Replay Workbench, Proxy, Bench, Monitor, Tools, CLI, and MCP.

CONNECTIONS

Connections and TraceGrid stay together

Create broker tabs, connect MQTT 3.1.1 / MQTT 5 clients, publish, subscribe, inspect payloads, resend, and keep trace rows in the same workspace.

PCAP

PCAP diagnostics point to source evidence

Analyze captures into packet grids, traffic summaries, diagnostic checks, findings, issue groups, replay candidates, and bounded reports.

REPLAY

Replay Workbench turns evidence into scenarios

Import a Replay Source, build a draft, dry-run it, execute against a target broker, verify expected MQTT responses, and keep report history.

PROXY

Proxy creates controlled failure

Inject latency, bandwidth limits, timeouts, and disconnects with cleanup and recovery verification instead of improvising network chaos.

BENCH

Bench explains capacity work

Configure client groups, subscriptions, scheduled messages, preflight checks, live charts, history comparison, and long-running evidence.

MONITOR

Monitor separates app pressure from broker behavior

Runtime metrics and structured product events show memory, file descriptors, process health, and support-ready diagnostic evidence.

CLI · MCP

CLI and MCP help agents without hiding the app

Agents discover capabilities, read schemas and resources, prepare masked confirmation plans, apply only after confirmation, and verify results.

Daily debugging path

Connect, capture, reproduce, prove.

The product is built around the path engineers already take during MQTT incidents, not around a generic SaaS dashboard.

01

Connect real clients

Open a broker tab, create MQTT client identities, test TCP/TLS/WebSocket/QUIC paths, subscribe, publish, and watch live trace rows.

02

Capture the evidence

Use TraceGrid and PCAP Analysis to preserve ordered packet rows, payload metadata, diagnostics, findings, and source references.

03

Reproduce the behavior

Turn trace or PCAP evidence into a Replay Draft, dry-run the plan, execute against a target broker, and compare expected responses.

04

Package the proof

Use Replay Reports, Benchmark reports, Monitor events, and CLI/MCP verification so another engineer can rerun the same reasoning.

Simple Free / Pro boundary

Free for local MQTT debugging. Pro for scale and deeper evidence.

Free keeps daily MQTT work usable. Pro expands limits for serious Replay, PCAP, Benchmark, Remote Runtime, and advanced MCP-assisted diagnosis.

Free

$0/forever

For local MQTT debugging and daily trace work.

  • MQTT 3.1.1 and MQTT 5 clients with TCP, TLS, WebSocket, QoS, retain, will messages, publish, subscribe, trace, Monitor, Tools, and basic exports
  • Connections, PCAP Analysis, Replay, Proxy, Monitor, and Tools for everyday local diagnosis with practical Free limits
  • Benchmark runs within Free duration limits, including preflight, runtime context, and local evidence handoff
Early annual

Pro Annual

$49.90/year
$99

For engineers and consultants who need larger workspaces, deeper evidence, Remote Runtime, and Pro MCP workflows.

  • Larger Broker, Proxy, Replay, and PCAP workspaces plus Remote Runtime access
  • Replay above Free step limits, advanced pace controls, longer Bench duration, and complete PCAP diagnostic details
  • Advanced MCP workflows for deep diagnosis, TLS analysis, benchmark comparison, PCAP root-cause reports, and report persistence
  • Annual access to Pro MCP workflows, Remote Runtime, and MQTT evidence features
Lifetime

Pro Lifetime

$199

For engineers who want one-time Pro access for the supported life of Mqttable.

  • Larger Broker, Proxy, Replay, and PCAP workspaces plus Remote Runtime access
  • Replay above Free step limits, advanced pace controls, longer Bench duration, and complete PCAP diagnostic details
  • Advanced MCP workflows for deep diagnosis, TLS analysis, benchmark comparison, PCAP root-cause reports, and report persistence
  • One-time Pro access for the supported life of Mqttable, no renewal

Direct answers

Built for MQTT engineers, not dashboard cosplay.

Short answers about the product boundaries this landing page now reflects.

Stop losing MQTT evidence

Open the workbench before the next trace goes cold.

Start with local Connections, TraceGrid, PCAP, Replay, Proxy, Bench, Monitor, and Tools. Bring in CLI or MCP only when the workflow needs automation.

Start the workspace