---
name: network-authority-and-exploit-review
description: Use when sensitive multiplayer or MMORPG client actions require static server-authority validation, trust-boundary, replay, abuse, or rate-limit analysis with no attack traffic.
version: 0.1.0
author: GameStudio-CodexKIT
license: MIT
compatibility:
  engines: [engine-agnostic, unity]
  versions: [any]
  platforms: [windows, linux]
metadata:
  studio:
    type: diagnostic
    lifecycle_stage: review
    risk_level: read-only
    packs: [cpp-lua-mmorpg]
    side_effects: none
    artifact: network-authority-review.json
    required_evidence: [trust-boundary-map, action-contracts, finding-list]
    owner: HoaTV Studio
    reviewer: null
    maturity: experimental
    last_reviewed: 2026-08-08
    provenance:
      derived_from: none
      patterns_from: [AGENTS.md safety invariants, static server-authority review]
      copied_text: none
---
# Network Authority and Exploit Review

## Overview
Review who may request, validate, authorize, rate-limit, persist, and broadcast sensitive game actions without performing offensive testing.

## When to use
Use for client-authoritative currency, inventory, combat, movement, reward, trade, admin, replay, duplication, or rate-limit concerns.

## When NOT to use
Do not use to send exploit payloads, bypass authentication, scan external systems, or replace an authorized penetration test.

## Required inputs and context discovery
Collect action list, trust boundaries, initiators, authoritative side, server validation, idempotency, replay protection, rate limits, persistence, audit logs, and supported threat assumptions.

## Safety and risk level
Read-only static review. No live traffic, credential use, service control, player-data access, or exploit proof is authorized.

## Workflow
1. Map clients, gateways, servers, databases, and administrative trust boundaries.
   Completion criterion: every sensitive transition has an owner.
2. Inventory sensitive actions and their authoritative side.
   Completion criterion: client-authoritative sensitive state is visible.
3. Review validation, idempotency, replay, sequencing, and rate-limit guards.
   Completion criterion: missing guards become discrete findings.
4. Trace persistence and broadcast behavior for abuse amplification.
   Completion criterion: exploit impact is bounded without attack execution.
5. Rank source fixes and safe regression simulations.
   Completion criterion: live exploit attempts remain BLOCKED.

## Evidence and output contract
Produce `network-authority-review.json` with trust boundaries, action contracts, findings, severity rationale, limitations, and defensive tests.

## Handoff contract
Record reviewed actions, authority decisions, missing guards, sensitive paths, untested live assumptions, and next defensive owner.

## Pitfalls and anti-rationalization
- Encryption does not replace server authority.
- A client UI restriction is not server validation.
- Do not prove impact on a live environment.
- Static review does not certify exploit absence.

## Verification checklist
- [ ] Trust boundaries are explicit.
- [ ] Sensitive actions have authority owners.
- [ ] Validation, replay, and rate limits are reviewed.
- [ ] Findings cite source contracts.
- [ ] No attack traffic was sent.

## References and scripts
Use the bundled [scripts/authority_review.py](scripts/authority_review.py) on normalized action declarations and project-specific server contracts.