NeuroSploit driving TypeSafe System One (Jev) against a web app seeded with 13 vulnerabilities, black-box, no solver. Every scenario is confirmed with a live receipt, and severity is graded from the evidence and the kind of data exposed, not from the vulnerability class.
The scenarios that needed a multi-step chain, re-run on the current build with TypeSafe off (A) and on (B). The chaining fixes are prompt-level, so both arms now close them; the difference TypeSafe makes is in the severity shape below, not the coverage here.
The eight full-surface scenarios (reflected / stored / SVG / DOM XSS, boolean-blind SQLi, login open-redirect) were confirmed in the prior full-surface run and were out of this focused re-run's agent scope; together the harness covers all 13.
The engagement also chained past the planted bugs into impact the target's own team can act on immediately, each proven end to end.
/login and rendered the /admin panel listing every user, a vertical privilege-escalation chain proven from a self-registered customer account./config.json and /app.js (CWE-200), a GraphQL authorization bypass with introspection enabled, and an authenticated RCE via a JS report-template upload.Same findings, graded by the two builds. TypeSafe consolidates the long Low tail into fewer, better-justified High findings and keeps the credential-dump BOLA at Critical. Severity is computed by the FIRST v3.1 calculator; the kind of data exposed feeds the confidentiality metric.
The object-level auth flaw on GET /api/v2/users/:id lets a self-registered customer token read
any user's full record, including the admin's plaintext password and live API key. The score is graded
from two axes: whether impact was demonstrated, and the kind of data that impact touched. A
credential and API-key exposure grants the confidentiality metric on its own, so the finding holds
Critical rather than being softened to a generic access-control note.
The number is computed by the deterministic calculator, not chosen by a model. TypeSafe's role is calibration: a `Choice` over confirmed / needs-review / rejected and a data-sensitivity `Score` that keeps a demonstrated secret exposure at its true weight while still deflating a class-inflated finding that shows no real impact. It never resurrects a rejected claim; the operator owns the final call.
claude-opus-4-8 via subscription, black-box,
--typesafe on, single-model vote, no pre-baked solver: the LLM discovered and confirmed every
finding live. Coverage is scored by class plus endpoint match against the target's 13-scenario ground truth;
a match is a confirmed receipt, not a graded proof. Severity is computed by the FIRST v3.1 calculator with an
evidence-and-data-type grading pass. Scope: one target, run at vote-n 1 (no cross-model
agreement), so this is one honest data point on one application, not a leaderboard. Every finding, its receipt
and the signed assurance manifest are in the run's artifacts.