Make the report prove itself.

Compile a strict evidence contract, compare vulnerable and patched runs, and block confident findings that cannot survive a counterfactual.

Teaching sample from the supplied PRUNE-crash claim. This tool validates submitted evidence; it does not execute libp2p or verify a live CVE.

Evidence bundle

Bundled sample

0 charsSHA-256 pending

TRIAGE DIAGNOSIS

NOT TESTED

No finding crosses the gate until its evidence contract compiles and its counterfactuals agree.

0/6

Vulnerable

waitingWAIT

Control

waitingWAIT

Patched

waitingWAIT

EXACT FAILURE LEDGER0 ERRORS

Validate a sample to see schema paths and evidence repairs.

Boundary: passing proves declared structure and internal evidence consistency, not that artifacts are genuine. Re-run supplied commands in isolated infrastructure and verify hashes before disclosure.

Confidence never enters the equation.

A report is a candidate. Reproduction needs a scoped version, repeated trigger outcome, stable failure fingerprint, a negative control, and a patched counterfactual. Miss one and the next action is evidence collection, not escalation.

Why a control run?

If an ordinary message also crashes the same environment, the crafted trigger is not isolated. Ambient instability can mimic exploitability.

Why test the patch?

A patched run with the same input is the strongest local counterfactual. If it still fails, the claimed version boundary or remediation is contradicted.

What does REPRODUCED still not prove?

It does not authenticate logs, prove exploit reachability on a live network, authorize disclosure, or replace expert review.

Super generates helpful tools and automates fact-checking across the internet proactively. If you enjoyed this tool, build your own with Super and share it with a friend.