webpage-to-markdown is a Chrome Manifest V3 extension that reads page content and writes Markdown files to disk. It runs with <all_urls> host access, so security reports are taken seriously — thank you for taking the time to send one.

Only the latest published release is supported. Fixes land on main and ship in the next release; there are no backports to earlier tags.

Before reporting, update to the latest version and confirm the issue still reproduces.

Do not open a public issue for a security problem.

Preferred channel — GitHub private vulnerability reporting. It keeps the report private until an advisory is published.

Fallback — email contact@quentinveys.be with [SECURITY] webpage-to-markdown in the subject.

• Affected version (manifest.json → version) and Chrome version
• A description of the impact — what an attacker gains, not just what misbehaves
• Reproduction steps, ideally with a minimal hostile page (HTML snippet is enough)
• Any logs, stack traces, or screenshots
• Whether you intend to disclose publicly, and on what timeline

You will get a first substantive reply within 15 days of your report. That reply states whether the issue is confirmed, its assessed severity, and the expected timeline for a fix.

This is a solo-maintained hobby project, not a funded product: the target above is best-effort, and there is no bug bounty. If a deadline slips you will be told why.

Coordinated disclosure. Please give up to 30 days from acknowledgement before disclosing publicly, so a fix can ship and users can update. A GitHub Security Advisory is published alongside the fix, and reporters are credited by name or handle unless they ask otherwise.

Research conducted in good faith under this policy — on your own browser profile and your own test pages — will not be pursued. Do not test against other people's data, do not run denial-of-service or automated scanning against third-party sites through the crawler, and do not access or exfiltrate data that is not yours.

• Code execution or privilege escalation from hostile page content reaching the service worker, side panel, popup, options page, or offscreen document
• Markdown or filename injection leading to path traversal or overwriting files outside the chosen download directory
• Leaking captured content, settings, or browsing history off the machine
• Bypasses of the extension pages' Content Security Policy
• Crawler behaviour that ignores its own URL-scheme, same-scope, or asset filters in a way that leaks credentials or reaches unintended hosts
• Insecure handling of chrome.storage.local data (settings, session state, crawl queue)

• Vulnerabilities in Chrome itself — report those to the Chrome VRP
• Findings that require an already-compromised browser profile, a malicious extension installed alongside this one, or physical device access
• Vulnerabilities in vendored third-party libraries — report upstream first (Readability, Turndown, turndown-plugin-gfm); see NOTICE. Tell us too if the extension amplifies the impact.
• Dev-only npm dependencies (the jest chain) — they never reach shipped code
• The breadth of the requested permissions, absent a concrete exploit. The rationale for each one is documented in the README.
• Missing hardening with no demonstrated impact (e.g. "no Subresource Integrity"), reports produced solely by an automated scanner, and social-engineering scenarios

Useful context when assessing a finding:

• No backend, no telemetry. The extension has no server. Captured content, settings, and history stay in chrome.storage.local and in the files you download.
• One outbound request path. Only the crawler performs network requests (js/crawl-engine.js), exclusively to the pages you asked it to crawl, with credentials: "omit" so cookies and auth headers are never attached.
• Zero production dependencies. Nothing is fetched from npm at runtime; the three third-party libraries are vendored, pinned, and modified only by upstream updates.
• Strict CSP. Extension pages run under script-src 'self'; object-src 'self' — no inline scripts, no remote code, no eval.
• Untrusted page HTML is parsed off the main context in the offscreen document (js/offscreen.js), which has no privileged extension APIs beyond messaging.
• Supply chain. GitHub Actions are pinned to commit SHAs, Dependabot watches the dev dependencies, and npm audit --omit=dev --audit-level=high gates every PR.
• Static analysis. CodeQL code scanning (default setup) runs on every pull request and on a weekly schedule.

Researchers who report valid issues are credited in the published advisory and in the release notes for the fix.