Roadmap

Current State (v1.0.0)

Completed Features

Core

• Multi-language parsing (TypeScript, JavaScript, Python, Rust, Markdown, SQL, C#, Django/Alembic)
• Config file parsing via config-file-parser.ts — .env, Dockerfile, docker-compose, K8s YAML, Terraform, GraphQL, GitHub Actions, GitLab CI, CloudFormation, OpenAPI, package.json, requirements.txt
• 37 extractors (code, data, infrastructure, config, security, Python runtime, Rust)
• Neo4j graph with 17 node types and 19 edge types
• Multi-project isolation via projectId
• File watcher with debounced reindex (Chokidar, 500ms) and automatic graph cleanup on file deletion
• Incremental scan (git diff-based)
• Confidence scoring on CALLS edges (same-file: 1.0, import-resolved: 0.95, global: 0.60)
• Env var deduplication — same var in .env and process.env = one node
• nomik init auto-generates .env with all config variables

MCP Server

• 21 tools (search, impact, trace, path, context, health, communities, flows, diff, rules, guard, rename, wiki, test-impact, audit, onboard, etc.)
• 9 resources (stats, health, files, communities, onboard, schema, projects, infrastructure, guard)
• 6 prompts (onboard, review-change, health-check, explain-module, migration-plan, infrastructure)
• Role-scoped access (dev, architect, security, pm)
• Sampling support (server-to-client LLM completions)

CLI (38 commands)

• Core: init, scan, scan:incremental, status, watch, query, recent
• Analysis: impact, explain, pr-impact, test-impact, rename, migrate, audit
• Architecture: rules (9 built-in + custom Cypher), guard, communities, flows, diff, onboard, wiki, badge, service-links, changelog
• Infrastructure: serve, dashboard (REST API with 14 endpoints), ci, doctor
• Setup: setup-cursor, setup-windsurf, setup-claude, setup-antigravity
• Projects: list, create, switch, delete, info

Parser

• Import-aware extraction (all extractors resolve receiver variables from actual imports, no hardcoded names)
• DB tracking: Prisma, Supabase, Knex, TypeORM, pg, mysql2, drizzle, Redis + SQL/C#/Python migration schema parsing
• Infrastructure: queues (Bull/BullMQ/Bee-Queue/Agenda/pg-boss), Kafka, RabbitMQ, NATS, SQS/SNS, Google Pub/Sub, metrics (Prometheus/OpenTelemetry), tracing (OpenTelemetry/dd-trace/Sentry), WebSockets (ws/Socket.io/uWebSockets), cron (node-cron/node-schedule/@nestjs/schedule)
• Config: Docker, CI/CD (GitHub Actions/GitLab CI), Terraform, CloudFormation, OpenAPI spec, GraphQL schema, dependencies (package.json/requirements.txt), dotenv — all wired via config-file-parser.ts
• Security: secret detection (AWS/GitHub/Stripe/Slack/JWT/private keys), feature flags (LaunchDarkly/Unleash/Flagsmith/Split/GrowthBook), env var tracking
• Python runtime: Redis, Celery tasks, Prometheus, OpenTelemetry, message brokers
• API: Express/Fastify/NestJS routes, gRPC, tRPC, GraphQL resolvers, Swagger/OpenAPI enrichment

Visualization

• 2D graph (Cytoscape.js) + 3D graph (Three.js / 3d-force-graph)
• Search with focused navigation (ranked results, prev/next)
• Filter panel, layout selector (Force/Tree/Radial/Circle), stats panel
• Level-of-detail (zoom-based), adaptive performance (4 quality tiers)
• Pagination (top 500 files by importance), dark theme, project selector

Quality

• Dead code detection (import-aware, excludes class methods, barrel re-exports, constructors)
• God file / god object detection (function count threshold, cross-file coupling threshold)
• Duplicate code detection (bodyHash SHA-256, excludes trivial stubs < 3 lines)
• Architecture rules engine (9 built-in + custom Cypher via .nomik/rules.yaml)
• PR impact analysis with risk scoring (LOW/MEDIUM/HIGH)
• GitHub bot for PR auto-comments with blast radius
• CI pipeline (nomik ci — scan, rules, guard, audit)
• Health badges for README (shields.io)

Next — Q2 2026

Future — Q3+ 2026