Skip to content

Latest commit

 

History

History
195 lines (137 loc) · 7.07 KB

File metadata and controls

195 lines (137 loc) · 7.07 KB

Burnzy System Standard v1

Every system, whether small or massive, must be truthful, precise, secure, auditable, and worthy of institutional respect.


Purpose

This is the engineering standard applied to every repository, application, protocol, dashboard, and service under this org. It is not aspirational — it is the minimum bar. Systems that do not meet this standard are either in active remediation or marked as pre-release.


The Six Gates

Every system must pass through six gates, in order. No gate can be skipped. Later gates do not compensate for earlier failures.

Gate 1 — Truth

Remove all fake claims, stale labels, and cosmetic confidence.

Check Criteria
README accuracy Every feature listed is demonstrably present in the codebase
Status labels No feature marked "Done" unless it compiles, runs, and is reachable
Dashboard honesty No vanity metrics, no misleading "healthy" indicators
Documentation alignment Docs describe what the system does, not what it might do
Error messages Truthful, actionable, never swallowed silently

Test: Could an auditor read your README and match every claim to running code within 30 minutes?


Gate 2 — Structure

Clean architecture, naming, ownership, and boundaries.

Check Criteria
Single responsibility Each file/module has one clear purpose
Naming Files, functions, and variables describe what they do without guessing
No dead code Every export is imported somewhere; every file is reachable
Config centralization One source of truth for configuration, not scattered env reads
Dependency hygiene No unused packages; pinned versions for critical deps
Type safety TypeScript strict mode, no any escape hatches without documented reason

Test: Could a new engineer navigate the codebase and find any feature within 5 minutes?


Gate 3 — Safety

Secrets, validation, permissions, rate limits, and failure modes.

Check Criteria
Secrets discipline No secrets in client bundles; server-only env vars never prefixed NEXT_PUBLIC_
Env validation Schema-based validation at startup; placeholder detection; paired var checks
Rate limiting Distributed (Redis/Upstash) for production; in-memory acceptable only for local dev
Auth boundaries Admin routes require authentication; public routes expose minimal data
Input validation All user input validated (Zod or equivalent) before processing
Failure handling Typed error codes; structured error responses; no raw stack traces in production
Network awareness Users see which network they're on before signing transactions

Test: Could a pentester find secrets in the browser bundle, bypass rate limits, or access admin data without credentials?


Gate 4 — Verification

Tests, build checks, health checks, and observability.

Check Criteria
Type checking tsc --noEmit passes with 0 errors
Unit tests Critical paths covered; tests pass in CI; no skipped tests without reason
Build verification npm run build exits 0; all routes compile
Verification pipeline Single command (npm run verify) runs typecheck → test → build
Structured logging JSON in production, human-readable in dev; log levels configurable
Operation timing Critical async operations (API calls, uploads, DB queries) timed and logged

Test: Can you run one command and know, within 60 seconds, whether the system is correct?


Gate 5 — Credibility

Docs aligned to code, dashboards aligned to reality, clean UX.

Check Criteria
README = reality Feature table matches running application
SETUP.md = working A new developer can go from clone to running in under 10 minutes
API docs = endpoints Every documented endpoint exists and behaves as described
Status endpoints Public: minimal (ok, version); authenticated: full diagnostics
UX honesty No buttons that don't work; no "coming soon" branded as "available"
Legal/compliance Required disclosures present if applicable; not fabricated

Test: Would a paying client, after reading your docs and running the setup guide, feel confident or confused?


Gate 6 — Expansion

Only after Gates 1–5 are satisfied.

Check Criteria
Extensibility New features can be added without rewriting existing ones
Modularity Core modules have stable APIs; changes don't cascade unpredictably
Migration path Database schema, API contracts, and config formats have upgrade paths
Performance budget Response times and bundle sizes measured, not assumed
Monitoring readiness Logs, metrics, and alerts can be wired in without code changes

Test: Could you add a major feature in one week without breaking something else?


Scoring

Each gate is scored on a 4-point scale:

Score Meaning
0 — Failing Gate requirements not met; system should not ship
1 — Partial Some checks pass but critical gaps remain
2 — Passing Meets minimum bar; no blocking issues
3 — Strong Exceeds standard; demonstrates engineering maturity

System rating = lowest gate score. A system with Gate 3 (Safety) at 0 and Gate 4 (Verification) at 3 is rated 0. The chain is only as strong as the weakest gate.

Rating Bands

Total Score (sum of 6 gates) Band Meaning
0–5 Pre-release Not shippable
6–9 Alpha Internal use only
10–13 Beta Limited external use with caveats
14–16 Production Deployable with confidence
17–18 Institutional Audit-ready, investor-grade

Application Checklist

Quick-reference for applying this standard to any repo:

[ ] README claims match running code
[ ] No dead features labeled as "Done"
[ ] TypeScript strict, 0 errors
[ ] Env validation runs at startup
[ ] No secrets in client bundle
[ ] Rate limiting on all public mutation endpoints
[ ] Admin routes require authentication
[ ] Test suite passes (npm test)
[ ] Build passes (npm run build)
[ ] Verification pipeline exists (npm run verify)
[ ] Structured logger integrated
[ ] SETUP.md produces a running system
[ ] Status endpoint reveals minimal data publicly
[ ] No overstated UI labels
[ ] Git clean — no uncommitted hardening debt

Current Ratings

Repository Truth Structure Safety Verification Credibility Expansion Total Band
solana-launcher 2 2 2 2 2 2 12 Beta
donk-ai 2 2 2 2 2 1 11 Beta

Target: Both repos to Production (14+) by next audit cycle.


The Operating Question

Before shipping anything, answer these without smoke:

  1. What is this system for?
  2. What is real right now?
  3. What is missing?
  4. What can fail?
  5. Who controls what?
  6. What proves correctness?
  7. What proves readiness?
  8. How does it evolve without breaking?

If every answer is clean and verifiable, the system is worthy.


Burnzy System Standard v1 — March 2026