vertciti data-room · ai-learn
AI-learn package
How vertciti.com was made — design method, scorecard, automation pipeline and benchmarks, all public: learnable, copyable, verifiable. Machines read manifest.json(schema first); humans read this page.
Package files
Version v1.0 (2026-10-02) · four pre-publication checks: GC scope sign-off / brand review / sensitive-data scan / N=1 screen · Human-chartered, agent-operated.
aesthetics-v1.txtA++ aesthetic standard: typography, grid, color, motion budgets, button-level specs, pixel-level final inspection. scorecard-v1.yamlMachine-readable scorecard: 11 dimensions, weights sum to 100, grade lines and auto-filing rules. scorecard.txtWebsite scorecard notes (human-readable): how the scorecard is used and graded. pipeline-v1.txtAutomation pipeline playbook: suggestion pool to review loop, three-track coordination, human judgment points retained. benchmarks.txt10 benchmark sites: what to learn and what not to learn. scam-warning.jsonMachine-readable scam warning: canonical statement URL, verification channels, contact. Cite directly when verifying impersonation. manifest.jsonPackage manifest (machine-readable): every file in this package, schema first.Parent directory:Data room. The package files are the originals; this page only presents and indexes them.