DevPlayground Curriculum & Execution Plan

Overview

DevPlayground delivers a production-shaped coding playground for interview preparation. It is local-first, docker-compose driven, and instrumented for observability and performance feedback. Tracks progress from foundations (DSA) through applied systems (Cloud/DevOps), with machine coding projects that mimic real interviews.

Tracks & Milestones

• DSA (tracks/dsa/)
  • Master streaming/heap patterns, emit ops-per-second metrics, achieve >100k ops/sec benchmark.
• Machine Coding (tracks/machine-coding/)
  • Build REST services with persistence, observability, chaos toggles, and perf gates.
• LLD (tracks/lld/)
  • Apply Strategy/Factory patterns with config-driven wiring and high coverage.
• Patterns (tracks/patterns/)
  • Replace conditional complexity with Strategy chains, provide golden tests and telemetry.
• System Design (tracks/system-design/)
  • Produce rubric-aligned design docs with capacity planning, tradeoffs, and diagrams.
• Performance & Observability (tracks/perf-obs/)
  • Instrument Prometheus/Grafana, define SLOs, automate perf regressions.
• Cloud & DevOps (tracks/cloud-devops/)
  • Ensure local docker spin-up, CI enforcement, and secure grading environment.

Promotion requires hitting track-specific gates: correctness >= 90%, perf gates met (p95≤80ms, error≤0.5%), and complete observability artifacts.

Seed Challenges

• dsa-streaming-median: Streaming median with memory/time caps, deterministic benchmark harness.
• mc-url-shortener: URL shortener with TTL, rate limiting, idempotency, and observability.
• patterns-strategy-discount: Strategy-based discount engine with golden tests.
• hld-realtime-chat: System design doc covering presence, ordering, scaling, and risks.

Execution Loop

1. Plan: Derive curriculum milestones & challenge specs.
2. Scaffold: Generate monorepo structure, infra, and templates.
3. Author: Implement challenge contracts, tests, load scripts, dashboards.
4. Judge: Execute deterministic tests (make test) and load (make load).
5. Report: Produce HTML reports with scores, traces, and next steps.

Local Setup

make run   # docker compose up -d --build
make test  # run pytest public suite (per language harnesses)
make load  # execute k6 scenarios and parse results
make report CH=mc-url-shortener  # generate scorecard + HTML report

• Primary DB: SQLite (WAL). Optional Redis toggle via .env.
• Observability: Prometheus, Grafana, Loki, OpenTelemetry traces.
• Ports (defaults): app 8080, Prometheus 9090, Grafana 3000, Loki 3100.

Coach vs Judge

Set MODE=coach for guided hints, MODE=judge for strict grading. Reports live under reports/<challenge>/<timestamp>/.

Security & Fairness

• Grading runs sandboxed with no network egress unless explicitly whitelisted.
• Chaos toggles simulate packet loss, DB outages, and rate limit spikes.

Next Steps

• Populate templates and challenge scaffolds.
• Implement grading/reporting scripts with rubric weights.
• Generate synthetic report demonstrating scoring pipeline.