On this page
If you are looking for a Shortest alternative because the project has gone quiet, the right replacement depends on what you used it for. For natural-language test files, Playwright Test Agents or Midscene.js are closest. For checking that an AI agent's change actually works, Reticle verifies it from network, app state and console and replays it with no model.
Is Shortest still maintained?
Shortest looks dormant. Its last npm release was v0.4.9 on 2025-04-01, and its last substantive commit was on 2025-08-18; a housekeeping commit landed on 2026-05-21, and the repository is not archived (GitHub). We read that as effectively unmaintained, not shut down.
What Shortest did
Shortest's description is "QA via natural language AI tests" (GitHub, shortest.com). You pass plain-English strings to shortest(), and Claude, through the Anthropic API, drives a Playwright browser to carry them out. It runs headless in CI with the API key as a secret. It is MIT-licensed with about 5,700 stars on 2026-10-25, free apart from Anthropic usage, and Anthropic is the only model provider in its configuration. No MCP integration was found.
What Reticle does
Reticle checks an AI coding agent's change from inside the running web app and returns a verdict with evidence. It reads the network calls, the application state and the console, and on React and Svelte it names the file:line to fix.
Your agent calls it through MCP while it works. Flows it drives are saved and replay deterministically, with no model in the loop. It also works on sites you don't own: it drives the page in its own browser and injects its SDK, reading network, console, DOM and routing; app state and file:line need the SDK in your own build. The SDK is open source (Apache-2.0).
Shortest vs Reticle at a glance
| Shortest | Reticle | |
|---|---|---|
| Status | Dormant since 2025 | Active |
| Tests | Plain-English strings in code | Flows your agent drives, saved and replayed |
| Who runs a test | Claude drives Playwright each run | Deterministic replay, no model |
| What a test reads | What Claude sees on the page | Network, app state, console and DOM after each step |
| Model provider | Anthropic only | None needed for replay |
| Licence | MIT | SDK Apache-2.0; server FSL-1.1 |
What to do with your Shortest tests
- Keep them if they pass. MIT code does not stop working because releases stopped, but you carry the maintenance and API changes.
- Plain-English tests in your repo: move to Playwright Test Agents for generated specs, or Midscene.js for natural-language steps.
- Checking an agent's change before merge: that is the job Reticle does. Your agent drives the flow once, Reticle saves it, and it replays on every change without a model.
Where Reticle is the better choice
- No model on every run. Re-verifying four flows costs about 480 tokens, about 251 times less than a model re-driving them, and replaying a flow eight times gave one verdict all eight times.
- Failures under a correct screen. A double-submit or a write that never persists fails, because Reticle checks requests and state.
- Active maintenance. Releases continue in the public repository.
On 88 seeded bugs, Reticle caught 85 with no false alarms (scorecard).
Try Reticle
Paste one line to your coding agent, and it installs and wires the SDK in your app:
Follow https://raw.githubusercontent.com/reticlehq/reticle/main/SKILL.md
The wider list is on the alternatives page.
FAQ
Is Shortest still maintained?
Shortest appears to be dormant. Its last npm release was v0.4.9 on 1 April 2025 and its last substantive commit was in August 2025, though the repository is not archived and saw a housekeeping commit in May 2026. Existing MIT-licensed tests keep working, but you should plan for maintaining them yourself.
What is the best alternative to Shortest?
The best alternative to Shortest depends on what you used it for. For natural-language end-to-end tests, Playwright Test Agents and Midscene.js are the closest. For verifying that an AI coding agent's change works, Reticle checks network, app state and console and replays saved flows with no model on each run.
How did Shortest work?
Shortest worked by passing plain-English test descriptions to a shortest() function. Claude, through the Anthropic API, then drove a Playwright browser to carry out each test, and the suite could run headless in CI with the API key as a secret. It was MIT-licensed and free apart from model usage.
Why Reticle
Your agent says it works. Reticle checks.
- Checks each change inside the running app: network, state and console, not a screenshot.
- Returns a pass or fail with the evidence and the file:line to fix.
- Open-source SDK on your machine, or the platform on every preview and PR.
No credit card required.