Project Genome automatically scans your repository and detects your tech stack, coding patterns, testing conventions, and deployment setup. Every expert agent decision is informed by your project's DNA.
📦 package.json
⚙️ tsconfig.json
🐳 docker-compose.yml
📋 .eslintrc.json
📁 src/
└── index.ts
📁 tests/
└── api.test.ts
🔧 vitest.config.ts
DETECTED TRAITS
Genome traits come from multiple sources, each with different confidence levels. The richest knowledge comes from expert observations — patterns discovered during real work.
Scanned from your repo files — package.json, configs, file structure. High confidence, always current.
Defined by your team. Organizational decisions and domain context that can't be auto-detected.
Shared from global or group-level genome. New projects inherit your organization's standards automatically.
Discovered from completed job artifacts using pattern matching. Builds knowledge over time.
Captured by expert agents during execution. Quality-gated, then presented for your approval before becoming a trait.
Compare two projects' convention sets. Shared traits are candidates for global promotion. Conflicting traits highlight divergence that may cause agent inconsistency.
The genome scanner reads your repository files and detects the stack, frameworks, testing tools, styling approach, architecture patterns, and naming conventions. These traits filter the expert catalog so only relevant agents are offered during job decomposition.
Related Features
Request access to see auto-detected conventions for your repositories.
Request Early Access