Privacy Policy
1. Who we are
AIssue is an error monitoring service operated by Sousa Dev, Unipessoal Lda, a company registered in Portugal with NIPC 519697430 ("Sousa Dev", "we", "us").
This policy explains what personal data we process when you visit our website, create an account and use AIssue, and what we do with the error data your applications send to us. Our contact details are at the end of this page.
2. Our two roles
- Controller. For your account, your use of the website and dashboard, and security records, Sousa Dev decides how and why the data is processed.
- Processor. For the error data your applications send to AIssue (which may contain personal data about your users), you are the controller and Sousa Dev processes it only on your behalf and on your instructions, as described in the data processing section of our Terms of Service. If you are a user of an application that uses AIssue, please contact that application's operator about your data.
3. What we collect
Account data
Your username, email address and password (stored only as a salted hash). If you sign in with Google or GitHub, we receive your name, email address and account identifier from that provider; we do not receive your password there. We also store your preferences (such as dark mode), your plan, your monthly usage (events received and AI analyses used) and, if you came from one of our pages (such as the free dependency check or the pricing page), which page that was. We note that page in your sign-in session when you open the sign-in or sign-up page, not in a separate cookie.
Domains, settings and API tokens
The domains (projects) you create, their notification email addresses and settings. Account API tokens are stored only as a hash; we show a token once, when you create it.
Error data sent by your applications
When your application reports an error, we receive and store: the exception and stack trace, the HTTP status, the request path and method, request parameters, a limited set of request headers (such as user agent and referrer), the release, environment, platform and SDK name, the time, and, only if your application sends them, a user identifier and email address.
Before anything is stored, values that look like secrets (passwords, tokens, API keys, cookies, authorization headers and similar) are replaced with [Filtered]. Tracebacks, request data, headers and user identifiers are encrypted in our database.
Source code from GitHub (only if you connect a repository)
If you connect a GitHub repository, AIssue uses its GitHub App to read the files, functions and recent changes related to an error when it analyses it. The app can only read code (contents and metadata) and, to open GitHub issues, write issues; it cannot change your code. We store the repository name and identifiers; we do not store your GitHub password or a GitHub user token. File contents are cached temporarily (up to 6 hours). The code excerpts included in an analysis are stored, encrypted, with that analysis.
GitHub issues (only if you allow them)
Opening GitHub issues is off unless you allow it for a repository. If you do, and you open a GitHub issue from AIssue (or turn on automatic GitHub issues for a private repository), AIssue posts it to your repository as the AIssue app. It contains the issue title and location, occurrence counts and dates, the release and environment, the AI analysis and the latest stack trace, with secrets and email addresses redacted. Request data, headers and user identifiers are never posted. We store the GitHub issue's number and link. GitHub then stores and shows the issue under its own terms; if the repository is public, anyone can read it.
Dependency checks (only if you turn them on)
You can add GitHub repositories to your account with or without linking them to a domain. If you turn on dependency checks for a repository, AIssue reads its lockfiles and pinned requirements (read-only) when the default branch changes, checking for a new commit at most once a day, and stores the list of files checked, the package names and versions it uses, the vulnerable package versions it finds and the matching public advisories. To find vulnerabilities, AIssue sends package names, ecosystems and versions to OSV.dev; nothing else from your repository is sent. Package versions are looked up once for all repositories that use them, without saying which repository uses them. You can also add a project without connecting GitHub by pasting or uploading its lockfile: AIssue reads it once, stores the package names and versions it lists (not the file itself) and never re-reads it; you update it yourself. When new vulnerabilities are found, AIssue emails the domain's alert addresses (or, for a repository without a domain, your account email) with the advisories, fix steps and a prompt for a coding agent.
AI-BOM and SBOM
For each repository, AIssue builds a software bill of materials from the package names and versions it already stores for dependency checks. To add each package's declared license, AIssue sends package names, ecosystems and versions to deps.dev (Google's Open Source Insights); nothing else is sent. The AI bill of materials stores the answers you or your coding agent give about the repository's AI (such as its purpose, the people it affects, the data it uses and a contact for AI incidents) and the AI components found. If you turn on the code scan for a GitHub repository (off unless you turn it on), AIssue reads its code with the same read-only access and stores only what it finds: file paths and line numbers, model names, AI service addresses, the names (never the values) of AI API key variables, and the names of MCP servers. It does not store the code. Turning the scan off stops it; removing the repository deletes all of this.
Free tools (no account needed)
The free dependency check on our website asks for a lockfile and an email address. AIssue reads the file once and stores the file name and the package names and versions it lists (not the file itself), the email address, a one-way hash of your IP address (to limit how often the check can be used) and whether you asked for product updates. Package names, ecosystems and versions are sent to OSV.dev to find vulnerabilities, and the report is emailed to the address you gave. We use the address to send that report, and to send product updates only if you ticked that box.
The free GitHub repository check works the same way for a public repository you name: AIssue reads its lockfiles and stores the repository name, the commit checked, the lockfile paths and the package names and versions, with your email address, the hash of your IP address and your product-updates choice.
The free licence check and SBOM summary work like the dependency check, and also store the kind of project you chose and, where your lockfile records them, the licence each package declares and which packages are direct dependencies. Package names, ecosystems and versions are sent to deps.dev to look up licences and, for the SBOM summary, to OSV.dev to find vulnerabilities.
The EU AI Act risk check does not store your answers to show your result: they are kept in the result link itself, signed so they cannot be changed, and the link works for 30 days. Anyone you share the link with can see the answers in it; the description of your system is never put in the link, and is kept only in your browser session. If you ask for the checklist by email, we store your email address, your answers, the result, a hash of your IP address and your product-updates choice. If you send a request to talk to us, your message, email address and answers are emailed to our team and not otherwise stored by the tool.
The MCP config audit runs in your browser: the configuration you paste is not sent to us. If you ask for the report by email, we store your email address, the names of your MCP servers, the ids of the checks that matched and the grade (never the configuration or any value from it), a hash of your IP address and your product-updates choice, for 90 days.
The advisory digest stores your email address, the package names you follow (not versions), your settings and your product-updates choice, until you unsubscribe; nothing is sent until you confirm the address. Unsubscribing deletes them and keeps only a one-way hash of the address, so we never email it again unless you sign up again. We keep a record of which advisories each digest contained for 90 days, and delete unconfirmed sign-ups after 7 days. Package names are sent to OSV.dev to find advisories, without your address.
If you turn on a README badge for a repository, anyone with its address can see the number of known vulnerabilities in that repository by severity. The address uses a random token, never the repository's name, and never shows package names. Turning the badge off stops it.
For the public package and advisory pages, we count visits to a package version page, without storing who visited (a hashed visitor key kept for one day prevents counting the same visitor twice), and use a hash of your IP address for a few minutes to limit automated requests.
AI analyses
For each issue, the explanation, root cause, suggested fixes and the text sent to and received from the AI model are stored, encrypted, with the issue.
Security and technical data
Our servers log requests (IP address, requested URL, time and browser user agent). Sign-in attempts are recorded with the IP address and user agent to protect accounts against password-guessing attacks. We also count events per domain per day to show statistics and to apply rate limits.
Website analytics
To understand how our website and dashboard are used and to improve them, we use Umami, a privacy-focused analytics tool that Sousa Dev hosts itself, so no visit data goes to a third party. It uses no cookies and does not store IP addresses. It records the pages you visit, the referring site, your browser, operating system, device type and approximate country, and interactions such as button clicks (for example, which sign-up button was used). Links that carry one-time keys (such as password reset and email verification links) and search terms are removed from the addresses we record. When you are signed in, these visits are linked to an internal account number, never to your name or email address.
You can turn this off at any time in Your analytics choices at the end of this page. If your browser sends a Do Not Track or Global Privacy Control signal, we record nothing at all. Our public status pages, and the status widgets our customers embed on their own sites, are never tracked.
Session recordings and heatmaps (only with your consent). If you click "Allow" when we ask, we also record how you use our pages (page layout, scrolling, mouse movement and clicks) to produce session replays and heatmaps. Everything typed into forms is hidden, text shown in the dashboard is replaced with placeholders, and areas that show API tokens or payment details are not recorded at all. Pages whose links carry one-time keys are never recorded. When you are signed in, a recording is linked to your internal account number, so it is not anonymous. If you choose "No thanks", or do nothing, nothing is recorded. You can withdraw your consent at any time in Your analytics choices.
Payments (paid plans)
Payments are handled by Paddle.com, our reseller and the Merchant of Record for all our orders. Paddle collects your name, email address, billing address, payment details and, for businesses, VAT ID, and decides how to process them as an independent controller under its own privacy policy. Card details never reach AIssue. We receive and store Paddle's customer and subscription identifiers, your plan, billing period, currency, payment status and the amounts and dates of your payments, and we record which version of our terms you accepted at checkout.
Your own AI key (only if you add one)
On paid plans you can add a key for an AI provider of your choice. The key is stored encrypted and is never shown again. Analyses then send the error data and code excerpts described under "AI analyses" to that provider, under your own agreement with it.
4. Why we use it, and our legal bases
- To provide AIssue (accounts, receiving and grouping errors, AI analysis, email alerts, the dashboard and API): performance of our contract with you (GDPR Art. 6(1)(b)).
- To keep the service secure (sign-in protection, rate limits, abuse prevention, server logs): our legitimate interest in protecting the service and its users (Art. 6(1)(f)).
- To send service emails (error alerts you configured, account and security messages): performance of our contract.
- To keep aggregate statistics (numbers of errors received per day): our legitimate interest in understanding how the service is used. These counts contain no personal data.
- To bill paid plans and enforce plan limits (usage counts, usage emails, plan change notices): performance of our contract.
- To understand how our website and dashboard are used (cookieless website analytics, described in section 3): our legitimate interest in improving our website and product. You can object at any time (section 9).
- To record sessions and produce heatmaps (described in section 3): your consent (Art. 6(1)(a)), which you can withdraw at any time without affecting what was recorded before.
- To take payments (loading Paddle's checkout on the pages where you buy a plan or credits): performance of our contract with you, or steps you ask us to take before entering into it.
- To understand where accounts come from (the page an account signed up from): our legitimate interest in improving our website and product.
- To meet legal obligations, for example accounting and responding to lawful requests (Art. 6(1)(c)).
We do not sell personal data, and we do not use it for advertising.
5. Who receives data
We use the following service providers, who process data only on our instructions:
| Provider | Purpose | Data involved |
|---|---|---|
| Hosting provider | Runs our servers and database | All data described above |
| Anthropic (AI model provider) | Produces the AI analysis of an issue | The error (stack trace, HTTP status, request path and method, and request parameters with secrets removed) and, if a repository is connected, the related code excerpts. Request headers and user identifiers (user ID, email) are not sent. |
| Paddle.com (reseller and Merchant of Record) | Sells paid plans and credit packs, collects payment and taxes, issues invoices | Your name, email, billing address, payment details and VAT ID (collected by Paddle at checkout as an independent controller), and our customer and subscription identifiers |
| The AI provider of your choice (only if you add your own key) | Produces AI analyses with your key | The same error data and code excerpts as for Anthropic above, under your own agreement with that provider |
| Email delivery provider | Sends alert, account and security emails | Recipient address and email content (alert emails contain the issue and stack trace, not request data or user details) |
| GitHub | Only if you connect a repository or sign in with GitHub: lets AIssue read the repository's code, opens GitHub issues if you allow it, and signs you in | Repository and account identifiers; for GitHub issues, the content described in section 3 |
| Only if you sign in with Google | Account identifier, name and email | |
| deps.dev (Google Open Source Insights) | Declared licenses of package versions, for the SBOM and the free licence check; whether a version is public, for the public package pages | Package names, ecosystems and versions |
| OSV.dev (run by Google's open source security team) | Only if you turn on dependency checks: finds known vulnerabilities in package versions | Package names, ecosystems and versions from your lockfiles |
| Paddle's content delivery network | Delivers Paddle's checkout script, only on the pages where you buy a plan or credits | Your IP address and browser details, when your browser loads the script |
Our fonts, icons and scripts are served from our own servers; no other page asks your browser to contact a third party. Under Anthropic's commercial terms, data sent through its API is not used to train its models. If we add or replace an AI provider, we will update this list before using it. We may also disclose data when required by law.
6. International transfers
Some providers, such as Anthropic, GitHub and Google, are based in the United States. Where personal data is transferred outside the European Economic Area, the transfer relies on an adequacy decision (such as the EU-US Data Privacy Framework, for certified companies) or on the European Commission's Standard Contractual Clauses.
7. How long we keep data
- Account data: while your account exists. When you ask us to delete your account, we delete it and its domains, errors, issues and analyses.
- Error data and analyses: for the retention period of your plan, as shown on the pricing page (event details older than that are deleted daily; issues, statistics and the events that carry an AI analysis are kept), or until you delete or archive them, or delete the domain or your account, if that is sooner. Archiving a domain's errors permanently deletes the errors received until then, with their stack traces, request data and AI analyses, and the issues not seen since.
- Error statistics: the daily counts of errors received are kept after errors are archived or deleted, and after a domain or account is deleted. They contain no personal data.
- Cached repository files: up to 6 hours.
- Repositories and dependency check results: while the repository stays in your account, with its AI-BOM answers and code scan findings. Unlinking it from a domain keeps it in your list; removing it from the Security page (or deleting your account) deletes its package list, alerts and AI-BOM.
- Free tool requests (dependency, GitHub repository and licence checks, SBOM summaries, emailed EU AI Act checklists and MCP audit reports): 90 days, then deleted. If you asked for product updates, we keep only your email address and that choice, until you ask us to stop. Records used to limit how often the free tools can be used: 30 days.
- Advisory digest: until you unsubscribe (then only a hash of the address); the list of advisories sent, 90 days; unconfirmed sign-ups, 7 days.
- GitHub issues: archiving or deleting errors in AIssue does not delete issues already opened on GitHub; delete those on GitHub.
- Billing records (payments, invoices issued through Paddle, the terms you accepted): for as long as accounting and tax law requires.
- Server logs: for a limited time, for security and troubleshooting.
- Website analytics: page views and events, up to 2 years; session recordings and heatmap data, up to 90 days.
- Sign-in records: while your account exists.
8. Security
We protect data with encryption in transit (HTTPS), encryption at rest for error details and AI analyses, removal of secrets before storage, hashed passwords and API tokens, GitHub access limited to one repository that can read code and write only issues, and protection against password-guessing. No system is perfectly secure; if a breach affects your personal data, we will notify you and the authorities as the law requires.
9. Your rights
Under the GDPR you have the right to access your personal data, to correct it, to have it deleted, to restrict or object to its processing, to data portability, and to withdraw any consent you gave. To exercise these rights, contact us using the details below; we will answer within one month.
Your right to object. Where we rely on our legitimate interest (section 4), you can object at any time to that processing, on grounds relating to your particular situation. For our website analytics, you do not need to give a reason: turn it off in Your analytics choices at the end of this page, or turn on Do Not Track or Global Privacy Control in your browser, and we stop recording your visits.
You can also lodge a complaint with the Portuguese data protection authority, the Comissão Nacional de Proteção de Dados (CNPD), or with the authority of the EU country where you live or work.
10. Cookies
We only use cookies that are necessary for the service to work: a session cookie that keeps you signed in (up to two weeks), which also remembers which of our pages sent you to sign up while you do so, and a security cookie that protects forms against cross-site request forgery. Your browser may also store small settings for our pages, such as a remembered preference, and your analytics choices (whether website analytics are off and whether you allowed session recordings), so that we respect them and do not ask again.
We do not use analytics or advertising cookies: our website analytics (Umami, described in section 3) works without cookies, and session recordings start only after you allow them. If that changes, we will update this policy and ask for your consent first.
11. Children
AIssue is a tool for software developers and is not intended for anyone under 16. We do not knowingly collect data from children.
12. Changes to this policy
We will update this policy when our practices change, and change the date at the top. If a change is significant, we will tell account holders by email before it takes effect.
Your analytics choices
These choices are saved in this browser only. Clearing your browser's site data resets them.
Your browser asks sites not to track you (Do Not Track or Global Privacy Control), so we record nothing at all, whatever you choose here.
Cookieless page views and clicks (section 3). You can object at any time.
Only with your consent. Typed text and dashboard text are always hidden.
Contact us
AIssue is a product of Sousa Dev, Unipessoal Lda, NIPC 519697430.
For questions about this page, your data or your account, contact us at:
Email: henrique@sousadev.com