GenXisDashboard

Evidence-bound verification gate

Check the claim.
Keep the boundary.

A named artifact. A declared check. A result you can inspect.

GenXis GavelTM is separate from the proposing agent or system. It checks only the named artifact you submit against a declared check, then returns a bounded decision and signed receipt. Unverified is never a pass.

GenXis Gavel™
Agents propose. GavelTM proves only within the declared check.

From install to first proof

01 / one command · one email · no password
npm install -g genxis-gavel

The verification path

NAMED ARTIFACT

You choose the artifact and declared check.

BOUNDED CHECK

A separate verifier evaluates only that evidence.

BOUNDED DECISION

Pass, denial, or unverified. Your policy decides what advances.

SIGNED RECEIPT

It records the binding and claim ceiling, not a broader guarantee.

Developer workflow

Inside the loop you already use.

  1. 01Install the public package.
  2. 02Confirm the API and credential.
  3. 03Submit the bounded artifact.
  4. 04Keep or verify the receipt.
# Install
npm i genxis-gavel

# Confirm readiness
npx gavel status --json

# Check one artifact
npx gavel check .\change.patch --json `
  --receipt-out .\gavel-receipt.json

Why we built it

LLM Confidence is cheap. Mistakes are not.

We built GavelTM from the frustration of watching agent work arrive with confident completion claims, unclear evidence, and too much human time spent finding failures after the fact.

We wanted a separate verification step inside the loop, not another model grading itself. GavelTM makes the submitted evidence and its limits explicit. It does not replace human review, CI policy, security engineering, or production controls.

Verified in the wild

311Certified catches this week

Rolling seven-day total from all distinct Most Annoying AI benchmark receipt roots plus the production catch ledger.This week · refreshes automatically

A launch allowance for reproduced findings.

No card is needed to begin. Pass, denied, malformed, duplicate, timed-out, and unverified outcomes are not billed as findings. A new account receives a $5 launch allowance, applied only if an eligible bounded finding is independently reproduced.

Launch allowance

$5 credit

Applied to eligible independently reproduced findings on each new account.

Pass / no catch

$0

Clean outcomes stay free, before and after the launch allowance.

Small certified catch

$0.10

A verifier-certified medium, low, informational, or minor finding.

Major certified catch

$0.50

A verifier-certified high, critical, major, or material finding.

What earns a 10¢ catch?

SMALL

A bounded defect with localized impact, certified at medium, low, informational, or minor severity by the verifier responsible for that domain.

  • Localized schema or contract mismatch
  • Bounded reliability or configuration drift
  • Low-impact regression with reproducible evidence
  • Domain-policy violation that does not cross a major-risk threshold

What earns a 50¢ catch?

MAJOR

A high, critical, major, or material defect with independently reproduced evidence and a receipt bound to the checked artifact.

  • Recognized credential or private-key exposure
  • Shell, code, or SQL injection predicates
  • Disabled TLS or authenticated plaintext transport
  • Unsafe deserialization, JWT bypass, weak secret hashing, or unsafe archive extraction

The current public deterministic security mesh contains high and critical predicates, so its certified findings enter the 50¢ tier. The 10¢ tier applies when an enabled domain verifier emits a certified medium, low, informational, or minor finding. These examples describe bounded predicates, not complete security coverage.

Malformed, duplicate, timed-out, denied, or unverified outcomes are not billed. The launch allowance applies only when a bounded finding is independently reproduced.

Claim ceiling

Precision includes knowing where proof stops.

GavelTM records only the bounded check named in its receipt. It makes failures and uncertainty explicit instead of stretching a result into a larger promise.

  • Complete application security
  • General reasoning or intelligence
  • Autonomous authorship
  • Correctness beyond the declared evidence

Catch boundaries

What does a bounded decision cover?

What is a generalized catch?

A reusable predicate that can apply across many repositories or domains: a recognized secret format, shell-backed execution, disabled TLS verification, or another precisely declared pattern.

General reach does not make it certified. The candidate must still be reproduced by its verifier and bound to the submitted evidence.

What is a verified domain catch?

A finding judged inside a declared domain contract: a frozen benchmark answer, schema, policy, regression command, test oracle, or other bounded source of truth.

The GavelTM decision covers only that domain predicate for that evidence. It does not automatically generalize outside the contract.

What does not get a positive GavelTM decision?

  • Model suspicion without independent reproduction
  • Missing, malformed, ambiguous, or out-of-scope evidence
  • Duplicate, timed-out, or failed verifier output
  • A claim broader than the verifier can prove
  • Any candidate for which no suitable verifier exists

Fail-closed answer: those outcomes may receive an explicit denied, invalid, or unverified status, but they do not become a certified catch, a pass, or a promotion. No evidence never means success.

Your first check

Check the namedevidence.