Hlido

kubefwd

Infrastructure · tested 2026-08-15 · re-test due 2026-11-15 · by the Hlido desk, not the vendor

In short: A mature CNCF bulk Kubernetes port-forwarder that now speaks MCP — the AI angle is a thin but well-executed graft onto a tool that already earned its trust.

7 PASS · 0 FAIL of 7 public-surface claims

Visit kubefwd → scorecard.json

Quick answer

kubefwd scores 80/100 (STEADY) on Hlido’s independent, hands-on test (reviewed 2026-08-15). STEADY (80) for a mature, honestly-marketed CNCF project (Landscape since 2018, Apache 2.0) with a clean public surface, multiple documented interfaces (TUI, 40+ endpoint REST API, drop-in MCP server), and clear multi-pl Pricing: Open source (free entry point documented).

Hlido Editor · 2026-08-15

kubefwd has been on the CNCF Landscape since 2018 and does one unglamorous thing extremely well: it bulk-forwards Kubernetes services to local loopback addresses, giving each service its own 127.x.x.x IP and writing the cluster hostname into /etc/hosts, so in-cluster connection strings (http://api:8080, redis-cli -h cache, mysql -h db) resolve locally without modification. Many databases on port 3306, many gateways on 443, no remapping, no environment-specific config. It watches Kubernetes events and reconnects on pod churn with exponential backoff. Apache 2.0, installable via brew, winget, apt or a release binary. The surface is clean and honest — nothing here overshoots. For the agentic-tooling question Hlido cares about, the relevant hook is the drop-in MCP server: Claude, Cursor and any MCP-capable assistant can connect, browse cluster services and forward what a task needs in plain language ('connect me to staging redis'), alongside a 40+ endpoint REST API and an interactive Bubble Tea TUI with live traffic sparklines and pod-log streaming. Crucially, the MCP server is presented as one interface among several rather than a bolted-on gimmick, and it sits on top of a tool whose core value is deterministic devops plumbing that predates the AI framing by years. The honest limits: it requires sudo (it writes /etc/hosts and binds low ports), so it runs with real local privilege; the AI-assistant surface is a small slice of what is fundamentally a platform-engineering utility; and Hlido assessed the marketing site only, not the running binary. Nothing on the captured surface makes a claim it cannot obviously back.

Why STEADY

STEADY (80) for a mature, honestly-marketed CNCF project (Landscape since 2018, Apache 2.0) with a clean public surface, multiple documented interfaces (TUI, 40+ endpoint REST API, drop-in MCP server), and clear multi-platform install paths, marked as low-medium confidence because the review is surface-only (the marketing site, not the running binary), the AI/MCP surface is a thin slice of a core devops utility, and the tool requires sudo to operate. Not VITAL because the agent-facing surface is one feature among many and nothing was hands-on tested.

Public-surface checklist

What we saw

1 screenshot captured by the Hlido engine during the reviewed run (run-3fa184f616ea72b2-kubefwd-com). Our own captures — not vendor marketing material.

kubefwd — run screenshot 1 (home.png)
home.png

What it does well

What it fails at

Best for

  • Platform and devops engineers who need many cluster services accessible locally exactly as they resolve in-cluster
  • Developers who want an MCP-drivable path for an AI assistant to reach staging/dev cluster services
  • Teams already in the Kubernetes/CNCF ecosystem wanting a well-established, Apache-2.0 local tool

Not recommended for

  • Non-technical users — this is a sudo-level command-line utility
  • Environments where root-level local tooling that edits /etc/hosts is disallowed by policy
  • Anyone wanting a managed or hosted service rather than a self-run binary

Pricing & access

Derived from Hlido-held evidence only (engine checklist + editorial text); quotes are verbatim from the scorecard; not vendor-supplied; re-derived daily. Verify current prices on the vendor's pricing page. Last verified 2026-08-15.

Related agents

Agent relevance

API CLI MCP Behavioral-testable

Agentic-Commerce Readiness 79/100 · COMMERCE-READY

Independent readiness for agent delegation & transaction. How it’s scored · check live

A drop-in MCP server (`kubefwd mcp`, added via `claude mcp add kubefwd`) lets Claude, Cursor and any MCP-capable assistant browse cluster services and establish forwards in natural language. A 40+ endpoint REST API also drives forwards programmatically from CI or editor tooling. The tool itself is a Go CLI run with sudo.

Agent-friendly score: 8/10

Evidence

scorecard.json · transparency passport · registry · methodology

More: compare agents · best of · developer tools · incident registry

Verdict by Hlido Editor, our automated editorial system · Method: public-surface-tier-2+editorial-narrative-v2 · Methodology version 2026.05 · Next review due 2026-11-15

How this page was produced. The scores, claim verdicts and evidence come from automated hands-on testing of the product’s public surface. The written analysis is drafted by an AI system, and pages publish without a person reviewing each one. Hlido publishes this record and answers for it — tell us if anything here is wrong and we will correct it.

Embed this trust badge

Hlido trust score

Live, always-current independent score — free to embed on your site or README. No vendor pays for placement.

Markdown

[![Hlido trust score](https://hlido.eu/badge/txn2-kubefwd.svg)](https://hlido.eu/check/?agent=txn2-kubefwd)

HTML

<a href="https://hlido.eu/check/?agent=txn2-kubefwd"><img src="https://hlido.eu/badge/txn2-kubefwd.svg" alt="Hlido trust score"></a>