Everdone
Everdone is an AI service platform for software engineering and QA, built by Mumbai-based Cointab. Connect GitHub and five services document, review, secure, profile and test your code. Usage-based pricing, 200 free units per service, no per-seat fees.
What is Everdone?
Everdone is a platform of AI-powered services for software engineering and QA, published by Cointab Software Private Limited in Mumbai, India. Its stated positioning is Work-as-a-Service: services that carry out defined tasks and hand back finished work, rather than a chatbot waiting to be prompted.
Five services are live and can be used separately or together. CodeDoc connects to GitHub, lets you pick repositories, folders or files, and documents the code file by file, producing organised folders, summaries, cross-links and global search; it re-documents only the files that changed, once a day, and never writes to the repository. Documentation can stay private, be shared with the team, or be published by link. CodeReview analyses a pull request, a branch or a set of files, ranks findings by severity, suggests fixes with code examples, and supports assignment, status tracking, full history and a verification re-review once the fix lands. CodeSecurity narrows the same loop to vulnerabilities that are genuinely exploitable rather than scanner noise, explaining the exploit and its impact. CodePerformance looks for hot paths, inefficient I/O, database misuse and scaling risks, then confirms the improvement iteratively. TestCase takes a BRD, a PRD, user stories, screenshots or free text, splits the work into independently testable scenarios, and writes test cases with preconditions, steps, test data and expected results — covering happy paths, validation failures, edge cases, error handling, role checks and security probes — exportable to Excel. Two further services, DBDoc and ChangeLog, are announced but not yet live.
There is nothing to install: access runs through a GitHub App and OAuth, with no token, CLI or local configuration. CodeDoc covers JavaScript, Python, TypeScript, Java, Go, PHP, Ruby and others, is framework-agnostic, and skips non-text files. Public demos on open-source projects — Next.js, FastAPI, React, Django REST Framework, Terraform, OWASP ZAP — show real output before signup. Billing follows usage rather than seats, team members are unlimited, and the platform also pitches itself as the context layer that makes agents like Cursor, Claude and Copilot work better on a documented repository.
What it does
- Document an entire GitHub repository file by file, in minutes
- Keep that documentation current automatically, re-documenting only the files changed each day
- Have AI review pull requests, branches or individual files for bugs, security and performance
- Detect exploitable vulnerabilities and confirm the fix holds with a verification re-review
- Spot performance bottlenecks, inefficient I/O and scaling risks, then verify the gains
- Turn a specification, a screenshot or a plain description into executable test cases
- Track findings end to end: assignment, status, history, audit trail and Excel export
When to use Everdone / When not to
A quick filter to help you decide if Everdone is the right fit.
When to use Everdone
- Engineering teams onboarding new developers who need an unfamiliar repository explained file by file
- Teams maintaining large or legacy codebases where context walks out the door when people leave
- QA teams still writing test cases by hand from specifications, screenshots or user stories
- Developers preparing a repository for AI coding agents such as Cursor, Claude or Copilot
- Distributed, fast-growing teams that want developers, QA and product on one tool without per-seat fees
When not to use Everdone
- Teams hosting code outside GitHub: no GitLab, Bitbucket or Azure DevOps integration is documented
- Anyone looking for a test execution engine: TestCase writes test cases, it never runs them
- Teams hoping to replace a test management platform such as TestRail or Qase
- Engineering organisations that need programmatic access: no public API documentation exists
- Buyers who require a published sub-processor list or a SOC 2 / ISO 27001 certificate
How to use Everdone
A typical end-to-end flow, from setup to results.
- Create an account at app.everdone.ai/signup — no credit card required
- Connect your GitHub account through the GitHub App and OAuth; nothing is installed locally
- Pick one of the five services: CodeDoc, CodeReview, CodeSecurity, CodePerformance or TestCase
- CodeDoc: select repositories, folders or files, then launch the documentation run
- CodeDoc: switch on daily automation so only changed files are re-documented, and share by link
- CodeReview, CodeSecurity or CodePerformance: select a pull request, a branch or files and start the review
- Sort the findings by severity, assign them, and follow their status through to resolution
- Run a verification re-review to confirm the fixes hold; only re-reviewed files are billed
- TestCase: drop in a specification document, screenshots or a free-text description and let the AI extract the scenarios
- Review, assign and filter the generated test cases, then export the suite to Excel
Pros & Cons
Pros
- No per-seat pricing: developers, QA and product managers all work from the same wallet
- 200 free units per service without a credit card, plus public demos on well-known open-source projects: real output before signup
- Low entry price at 0.05 USD per file or test case during early access, half the 0.10 USD list rate
- No contract, no minimum, no commitment, and nothing to install beyond a GitHub App and OAuth
- Your code is never used for model training, and the repository is read-only: nothing is written back
- Clear legal footing: you keep your content and your outputs (terms 5.1 and 5.2), with a DPA built into the privacy policy and a named, reachable DPO
- A verification re-review, rare among AI review tools, confirms the fix holds — and only re-reviewed files are billed
Cons
- GitHub is the only forge named on the site: no GitLab, no Bitbucket, no Azure DevOps
- No documented public API, no mobile application and no browser extension — the product is web-only
- No hosting country or region is named, no sub-processor list is published (written request only), and no security certification such as SOC 2 or ISO 27001 is displayed
- GDPR never appears in plain text on the pages, only in the meta keywords of the privacy policy
- Credits are non-refundable and non-transferable, expire twelve months after purchase, and there is no permanent free plan
- The 0.05 USD unit price is a 50% early access promotion; the reference rate is 0.10 USD
- Very young service (domain registered July 2025, first web archive capture November 2025) governed by Indian law, with arbitration and exclusive jurisdiction in Mumbai
Pricing & Plans
There is no permanent free plan, but the first 200 units of each of the five services are free and no credit card is required to start. Beyond that, Everdone bills on usage from a prepaid credit wallet: the list price is 0.10 USD per unit and the current early access rate is 0.05 USD, a 50% discount. A unit is one file documented (CodeDoc), one file reviewed per review (CodeReview, CodeSecurity, CodePerformance) or one test case generated (TestCase). There is no per-user fee, no contract and no minimum, and a re-review is charged only on the files actually reviewed again. All prices are exclusive of applicable taxes or duties, and unused credits expire twelve months after purchase, without refund or reinstatement.
- a single usage-based model
- identical across the five services
- CodeDoc — first 200 files documented free
- then 0.05 USD per file (early access
- 0.10 USD list price)
- CodeReview — first 200 files reviewed free
- then 0.05 USD per file per review
- CodeSecurity — first 200 files reviewed free
- then 0.05 USD per file per review
- CodePerformance — first 200 files reviewed free
- then 0.05 USD per file per review
- TestCase — first 200 test cases free
- then 0.05 USD per test case
- unlimited team members
- unlimited repositories
- team sharing and secure GitHub integration
Data, GDPR & hosting
A consolidated view of how Everdone handles your data.
GDPR overview
Everdone documents substantial GDPR machinery without ever claiming compliance in plain text: the word GDPR appears only in the meta keywords of the privacy page, in force since 15 October 2025. A data processing addendum is built into that policy at section 10 — the customer is controller, Cointab is processor acting on documented instructions, breach notification is promised without undue delay, and one written audit per year is allowed. Transfers out of the European Economic Area or the United Kingdom are covered by Standard Contractual Clauses or an equivalent lawful mechanism. Data subject rights (access, rectification, erasure, portability, objection, restriction) are offered subject to jurisdiction and exercised at privacy@everdone.ai after identity verification, and a DPO is named at the Mumbai address. Two gaps remain: no Article 27 EU representative is appointed, and no security certification is displayed.
Who owns the data?
Everdone's terms draw a clean line. Section 5.1 states that all data, code and materials you upload — your User Content — remain your sole property. Cointab Software Private Limited claims no ownership, only a non-exclusive, worldwide, royalty-free licence to process that content for the sole purpose of providing and improving the service. Section 5.2 gives you ownership of the Output: the documentation and analyses generated through Everdone are yours, subject to your compliance with the terms. Section 5.3 holds the other side of the line: the service itself, its software, algorithms and AI models remain Cointab's property. No third party is granted rights over your content.
Reuse rights
The outputs are yours to reuse, publish or redistribute without asking permission, provided you keep complying with the terms (section 5.2). CodeDoc documentation can stay private, be shared with the team, or be made public by link, and the publisher's own demo pages show generated documentation published openly. On its side, Cointab limits what it does with your data: it processes it to run the service, authenticate you, execute jobs and generate outputs, bill, support, monitor performance and security, prevent fraud, and develop and improve its products. It states that it does not sell or rent personal data and does not use customer data for advertising or profiling unrelated to service improvement; cookies are limited to session and aggregated audience measurement, with no advertising, behavioural or cross-site tracking. Your code is never used for model training. Sharing is confined to service providers (infrastructure, analytics, payments), legal obligations, corporate transfers and cases where you give explicit consent. Data collected covers the account, repository access tokens, execution logs, billing, support exchanges and technical data such as IP address, browser and session.
Data retention & training
Hosting summary
Everdone names no hosting country and no hosting region. The privacy policy calls the service global and states only that data may be stored or processed on servers located in jurisdictions other than the user's country of residence. Section 8 covers international transfers: where personal data leaves the European Economic Area, the United Kingdom or another region with transfer restrictions, Cointab says it puts appropriate safeguards in place, including Standard Contractual Clauses or another lawful mechanism. On security, the policy claims encryption in transit and at rest, access controls, environment separation, least-privilege access and continuous monitoring, but no certification — neither SOC 2 nor ISO 27001 — is displayed anywhere on the site. The publisher, Cointab Software Private Limited, operates from Mumbai, India, the only jurisdiction actually documented; the site itself is served behind Cloudflare on an anycast address (104.21.4.233), a CDN node that says nothing about where the data lives. In short: the legal framework for transfers is described, the physical location of the data is not.
Things to keep in mind
Risks and trade-offs to weigh before adopting Everdone.
- The advertised 0.05 USD is a 50% early access rate with no announced end date: the reference price is 0.10 USD, so budget on the higher figure
- Credits are prepaid, non-refundable, non-transferable and expire automatically after twelve months — buy only what you will actually consume
- The 200 free units per service are a one-off evaluation allowance, not a permanent free plan
- No hosting country is named, no sub-processor list is published, and no security certification is displayed: a data residency requirement cannot be satisfied from the public documentation
- Indian law, arbitration and exclusive jurisdiction in Mumbai; the service is provided as is, liability is capped at the fees paid over the previous twelve months, and features may be changed or removed at Cointab's discretion on a product barely a year old
- Unless you agree otherwise in writing, Cointab may identify you as a customer and use your name and logo for marketing and promotional purposes
- AI outputs may not be error-free and must not be relied on as professional advice: generated documentation, reviews and test cases still need a human reader, and a team that stops reading its own code loses the very understanding the tool was meant to restore
Setup & Integrations
Technical difficulty
Very low. There is nothing to install: no token, no CLI, no local configuration and no pipeline to build. You create an account, connect GitHub through a GitHub App and OAuth, then select repositories, folders or files in the interface, and the first results come back within minutes. Nothing has to be prompted or scripted, and TestCase accepts whatever input you already have — a specification document, screenshots or free text, even partial. The only real prerequisite is a GitHub account on which you are able to authorise a third-party application.
Deployment
Integrations
Behind Everdone
Fundraising
Social
Resources
All the official URLs gathered for verification and reference.
Alternatives
Tools that compete with or complement Everdone.
Frequently asked questions
Do I pay per user?
Is there anything to install?
Is my code used to train AI models?
Does Everdone modify my repository, and are private repositories supported?
Which languages does CodeDoc cover, and how does the documentation stay up to date?
Who can see the documentation Everdone produces?
What does a unit cost once the 200 free ones are used?
Who owns the output Everdone generates?
Do the wallet credits expire?
Is there a public API, a mobile app, or a stated hosting location?
Should you pick Everdone?
Everdone bundles five complementary services on one foundation: understand the code, review it, secure it, speed it up, and test what it does. The same GitHub connection feeds CodeDoc, CodeReview, CodeSecurity, CodePerformance and TestCase, and the whole team uses all five without anyone paying for a seat. Cost follows the code, not the headcount — unusual enough to matter for distributed or fast-growing teams.
Two details stand out. The verification re-review, rerunning an analysis to confirm a fix actually holds on security as much as on performance, is rare among AI review tools, and only the files looked at again are billed. The legal footing is also unusually clean for a young product: the terms leave your content and your generated output with you, the privacy policy carries a built-in data processing addendum, a DPO is named and reachable, and the site states flatly that your code is never used for model training.
The counterweights are real. GitHub is the only forge documented, there is no public API and no mobile app, and no hosting country or region is named anywhere. No sub-processor list is published and no SOC 2 or ISO 27001 certificate is displayed. The 0.05 USD unit price is an early access promotion at half the 0.10 USD reference rate, credits expire after twelve months without refund, and the 200 free units per service are an evaluation allowance, not a permanent free plan. The product itself is very young: the domain was registered in July 2025.
None of that has to be settled on trust. The 200 free units need no credit card, and the public demos on open-source projects show the real output before you sign up.
- Choosing a selection results in a full page refresh.
- Opens in a new window.