Sharajdev

AI Chrome extensions for developers

Eleven free Chrome extensions that apply AI to real engineering jobs: digesting repositories into token-budgeted prompt bundles, tracing bugs to code, auditing technical SEO, testing prompts, reviewing code, optimising queries, generating tests, assessing skill and answering questions from documentation.

AI Developer Tools

Free

litmus

System Prompt Tester

Test, analyze and version system prompts against the LLM you ship on — LLM-as-judge rubrics, deterministic tool/agent checks, variance and speed metrics. Local-first, BYOK, no backend.

Add to Chrome Explore litmus →
Free

RepoSpector

AI Code Review & Code Query Engine

AI-powered code review plus a natural-language query engine for any repository — get review feedback and ask questions about the codebase right on the page.

Add to Chrome Explore RepoSpector →
Free

speeDB

Grounded DB Query Optimiser

Scan any GitHub or GitLab repo for database queries and get grounded, output-identical rewrites re-verified against your actual codebase before you see them.

Add to Chrome Explore speeDB →
Free

EngGauge AI

Engineer Skill Assessment

Evidence-based engineer skill assessment from GitLab & GitHub review data. Your keys, your browser, no backend — every claim cites a specific pull or merge request.

Add to Chrome Explore EngGauge AI →
Free

QAtalyst

AI Test Case Generator

Turn any feature, user story or requirement into structured test cases — positive, negative, edge and boundary scenarios — in seconds. Stop writing QA test cases by hand.

Add to Chrome Explore QAtalyst →
Free

XHRScribe

AI API Test Generator

Capture live XHR and fetch network traffic and auto-generate runnable API tests with AI — payloads, headers and assertions included. No hand-written requests.

Add to Chrome Explore XHRScribe →
Free

Sherpa

Local Help Search

Crawl a documentation site, index it entirely in your browser, and answer questions from that index with source citations — zero servers, zero per-query cost.

Add to Chrome Explore Sherpa →
Free

Tracely

Trace Bugs to Code & Flows

Trace reported bugs to exact repository lines and broken user flows. Use local Ollama or your own API key to diagnose root causes with test verification.

Add to Chrome Explore Tracely →
Free

GitScribe

LLM Repo Digests & Code Maps

Turn GitHub or GitLab repos and PRs into token-budgeted LLM digests. Generate hallucination-checked architecture READMEs with zero tracking or telemetry.

Add to Chrome Explore GitScribe →

AI that works on your actual code and data

Most AI tools answer questions in a chat window that has none of your context. These extensions work from what's in your browser: the repository you're viewing, the traffic your app sends, the ticket on your board or the docs site you rely on. Several of them check the AI's answer before showing it. speeDB re-verifies every cited file and line against the repository, EngGauge AI cites a specific merge or pull request for every claim, and Sherpa answers with numbered citations and says when its index can't answer.

Which one should I use?

Local-first and bring your own key

Several of these tools have no backend. litmus runs in a side panel with your own API keys and no account, EngGauge AI keeps tokens and keys in your Chrome profile, and Sherpa bundles its embedding model inside the extension, so queries cost nothing. Before you install any extension, check its Chrome Web Store listing for the exact permissions and privacy details.

Going further

If you ship LLM features, read the guide on building LLM-as-judge rubrics for system prompt testing. For the non-AI side of testing, see the QA testing extensions and the API testing extensions.

More collections: QA testing extensions · API testing extensions · all extensions

Frequently asked questions

Are these AI extensions free?
Yes — they are free to install from the Chrome Web Store. Tools that call a hosted model use your own API key.
Do they send my code to a server?
Not to one run by Sharaj dev. Tools such as litmus, EngGauge AI and Sherpa have no backend. When you use your own AI provider key, the request goes to that provider.
Which models do they support?
It varies by tool. EngGauge AI, for example, works with local Chrome on-device AI or your own OpenAI, Anthropic or Gemini keys. Each tool page lists the details.
How do I stop AI suggestions from being made up?
Pick tools that check their output against your code: speeDB re-verifies every citation against the repository, EngGauge AI links each claim to a PR, and Sherpa flags questions its index cannot answer.