NEWThe release platform your coding agent runs — installs in one command

Stop clicking dashboards. Tell your agent to ship it.

Feature flags, A/B experiments, translations, and a feedback→fix loop — operated end-to-end by your coding agent. Describe what you want; it writes the code, ships it, and watches the metrics. Free to start, installs in one command.

12s installflags · A/B · i18n0 config files∞ experiments
Primitives

Four primitives. One platform.

Scroll. The panel pins, the demo morphs.

killswitches · live
claude >
new_checkouterrors 0.04%
ai_recommendationskilled in 187ms
legacy_uploadererrors 0.12%
runtime config · live
claude >
{
  "max_uploads": 100,
  "model": "opus-4.1",
  "enable_v2": true,
  "region": "us-east-1"
}
experiment · checkout_v3
claude >
checkout_v3
LIVE · DAY 6
A3.21%control
B4.82%▲ +49.8% · sig 99.2%
auto-collected · last 7d
claude >
Activation
42.1%
+3.2%
D7 retention
68.4%
+1.1%
$ / user
$47.20
+$2.18
Error rate
0.04%
−0.1%
01 · KILLSWITCHES

Kill a bad feature before it kills you.

Every flag is a killswitch. Pull it manually, or set a guardrail — error rate, p95 latency, spend — and Shipeasy yanks it for you.

  • Globally disabled in under 200ms
  • Logged, audited, reversible
  • Auto-armed by guardrail thresholds
02 · CONFIGS

Tweak runtime behaviour without redeploying.

Typed runtime configs targeted by user, plan, geo, or anything you send. Edit from Claude, the dashboard, or a PR.

  • Versioned and auditable
  • Per-environment drafts + publish
  • Auto-revert on regression
03 · EXPERIMENTS

A/B/n tests that know when to stop.

Sequential stats mean you can peek whenever you want without inflating false positives. Auto-ramping, SRM detection, guardrails — Claude writes the wrapper.

  • Sequential + Bayesian engines
  • Auto-ramp from 1% to 100% on a guardrail
  • SRM and outlier detection
04 · METRICS

The metrics you need, collected for you.

Activation, retention, revenue, and error rate are instrumented from day one. Add custom metrics with one line of code.

  • No SQL required for the basics
  • Custom events with one line
  • Streamed live to the dashboard
The loop

From bug report to merged PR, while you sleep.

The feedback widget catches it, Shipeasy triages it, your agent fixes it.

Step 01 · catch

A user reports it

The in-app feedback widget and the see() error primitive turn a bug or feature request into a structured, deduplicated item in your dashboard.

Step 02 · triage

Shipeasy structures it

Each report lands with the failing code path, the consequence chain, and a priority — ready for an agent to act on, not a screenshot to decipher.

Step 03 · fix

Your agent ships the fix

A scheduled Claude Code routine pulls the open queue, writes the fix on its own branch, and opens a pull request that closes the issue.

Step 04 · verify

Roll out behind a flag

Ship the fix behind a gate, ramp it on a guardrail, and watch the error rate fall on the auto-collected metrics — all in the same platform.

Pricing

Free to start. Fair as you grow.

Free
$0/ forever
Side projects and solo work.
  • 3 feature flags · 1 running experiment
  • Unlimited killswitches
  • 1M events / day
  • MCP + Claude integration
Start free
Enterprise
Custom
Self-hosted, audit logs, SLAs.
  • Everything in Team
  • Self-hosted option
  • SOC 2 Type II
  • Dedicated support
Talk to us
FAQ

You probably want to know.

How does the MCP install actually work?

One command — claude mcp add shipeasy — registers four tools with Claude. We use your project context to wire experiments into the right files. No YAML.

Does this only work with Claude Code?

No. Shipeasy is an MCP server, so the core — flags, experiments, configs, and i18n — works in any MCP-capable agent (Claude Code, Cursor, Windsurf, and others) by talking to the agent you already use. The deepest automation — the slash-command skills and the scheduled feedback→fix→PR routine — runs on Claude Code today, with more harnesses on the roadmap. You're never locked into a tool you don't already use.

Hosted, or self-host?

Both. Hosted is the default. Self-hosted is a one-binary deployment on Enterprise with full parity and an air-gapped mode.

Which stats engine?

Sequential testing by default — peek whenever, without inflating false positives. Bayesian and frequentist are available per-experiment.

What if Claude makes a mistake?

Every action is reversible and logged. Killswitches and experiment changes go through a PR or a tool call you can undo.

Which frameworks?

JS/TS, Python, Go, Ruby, Rust, Swift. The SDK is a thin client; the heavy lifting happens server-side.

Stop guessing. Start shipping.

Install in 12 seconds. Your first experiment before your coffee is cold.