security review, from the shipped code.
I find and explain client-side security bugs in modern web applications — nonce and CSP laundering, DOM XSS, deep-link and redirect handlers, and pre-authentication token flows. I work from the code you actually ship, prove impact with reproducible evidence, and write reports your engineers can act on.
Independent research on authorized bug-bounty programs and engagements. Manual, scoped, and honest about the difference between a lead and a bug.
what I do
JavaScript / SPA review
Static analysis of your production bundles for source→sink paths, nonce-copy primitives, innerHTML/insertAdjacentHTML sinks, and Trusted Types gaps.
CSP & nonce assessment
Is the policy actually protecting anything? I measure the enforced policy and find where a trusted script does the injection for you.
Auth & 2FA flows
Pre-authentication token timing and session issuance — whether a first-factor session is honoured before the second factor.
Finding triage
A second pair of eyes before you report: is it real, in scope, what is the impact, and how do you prove it.
how an engagement runs
- 01scope
Written scope, allowed hosts, test accounts, rules of engagement.
- 02recon
Public assets first — bundles, headers, policies — then authenticated paths.
- 03analysis
Manual review plus small purpose-built tooling. Every claim tied to a line of code or a request.
- 04proof
A deterministic reproduction with the minimum impact needed to show the issue.
- 05report
Root cause, impact, evidence, remediation, and a plain-language summary.
case studies
- CSP policy & parsing analysis
Measured a printed policy against live headers and proved the comma semantics with controls — a scope memo instead of a wasted hunt.
- Nonce-laundering trace
Traced a server value into a nonce'd inline script, then proved the server sanitised it — a documented negative with a reusable scanner.
- Pre-2FA token timing
A safe, repeatable harness for testing whether a first-factor token is honoured before the second factor.
selected work
- where the deeplink work can make money — and the reporting trap that kills it
desk research on which programs pay for nonce-laundering XSS.
- server-side trace of the deeplink gadget
a real sink, a real mount, and a sanitiser that held — reported as a negative, with evidence.
- the security toolkit
a dependency-free scanner and fetcher for the nonce-laundering class.
what I will not do
- · test outside the agreed scope or against accounts I do not own
- · run automated or mass scanning, brute force, or denial-of-service testing
- · inflate a finding or report a “missing header” as a vulnerability
- · touch real user data
get in touch
Tell me what you are building and what worries you. I will reply with a scope and a fixed price for the first pass.