Skip to content
AXLOOP AI
Menu
ProductMCP observabilityArchitectureCompanyEarly access

Product

Instrument the edge. Then govern it.

Three layers, one loop: an open-source collector at the edge, a self-hosted fleet dashboard that makes sense of it, and a policy layer that turns what you can see into what you can control.

01
The collector

Zero-code instrumentation for any MCP client.

Ships open source. Stores locally, syncs encrypted batches to your aggregation node — nothing is routed through us.

Desktop client plugin

Drops in beside an existing MCP config. Every tool invocation becomes an OTel span with no code change from the user.

Mobile SDK — iOS + Android

Swift and Kotlin instrumentation for agents running on phones, a surface most observability tooling ignores.

IDE extension

VS Code and JetBrains. Captures the MCP configs engineers wire up themselves, including the ones nobody registered.

Embedded sidecar

A drop-in process for agents inside your own apps. Language-agnostic: Rust, Swift, Python, Node, Go.

What every span carries

tool.namewhich MCP tool was invoked
tool.paramsredacted at the edge, per policy
tool.latency_mswith upstream breakdown
tool.statussuccess, failure, timeout
tokens.in / outcost attribution per call
device.id / user.idfleet drilldown keys

02
The fleet dashboard

Aggregate hundreds of devices into one coherent view.

Deployed on your Kubernetes cluster, or a single Docker Compose file for smaller fleets.

FLEET VIEW · ILLUSTRATIVE SAMPLE DATA128,400 MCP CALLS / 24H
DEVICES312+18 this week
ERROR RATE1.8%3 tools degraded
SPEND ATTRIBUTED$9,140across 14 teams

TOOL CALL STREAM

  • browser.navigatechrome · a.chen512 ms
  • internal.crm_lookupsidecar-07187 ms
  • shell.execmbp-14 · shadow-mcp73 ms
  • vector.searchandroid-14 · s.iyer264 ms
  • fs.read_filembp-14 · a.chen42 ms
  • postgres.queryios-17 · r.patel318 ms

LATENCY DISTRIBUTION

p50 88 msp95 640 msp99 1310 ms
CROSS-DEVICE TRACEagent.plan→mcp.tool_call→http.upstream→db.query

03
Roadmap

All three phases in order.

01
In development · Open source

The Collector

A zero-code instrumentation layer that wraps an MCP client and captures every tool call: name, redacted parameters, latency, success or failure, token usage. Stores locally, syncs encrypted batches.

  • Desktop client plugin
  • iOS + Android SDK
  • VS Code & JetBrains extension
  • Sidecar for embedded agents
02
Next · Self-hosted

The Fleet Dashboard

A self-hosted web UI on Kubernetes or Docker that turns raw spans into a fleet view — with drilldown all the way to a single device.

  • Cross-device flamegraphs
  • Fleet rollups + per-user drilldown
  • p50 / p95 / p99 by client type
  • Anomaly detection · cost attribution
03
On the roadmap · Planned

Policy & Governance

Observability becomes control: rules that decide which tools may run, on which devices, for which teams — with an audit trail behind every decision.

  • Restrict tools on unmanaged devices
  • Rate alerts per tool, per device
  • Shadow MCP detection
  • Audit log with conversation replay

From visibility to control

Once every tool call is a span, policy is the natural next step.

The same telemetry decides which tools may run, where, and for whom.