Skip to content

Internal Developer Platform

Welcome to the IDP MVP documentation. Use the navigation above to explore the platform.

Getting Started

Core Concepts

Team Management

  • Team Management — Onboard a team: namespace, SecretStore, ArgoCD ApplicationSet, Grafana folder, DORA metrics
  • GitHub App Setup — Replace PAT with GitHub App for higher rate limits and per-installation scoping
  • Scaling Runbook — Small/Medium/Large tiers, scaling signals, implementation status

Engineering Intelligence

  • Product Vision — What the intelligence layer answers, and for whom
  • Architecture — How it is built, and an honest inventory of which data actually exists
  • Maturity Model — Ad Hoc → Standardised → Platform Enabled → AI Enabled → Autonomous
  • Scoring — The evidence contract: no score without evidence, no number without data
  • Integrations — Every collector, its source, and what happens when that source is down
  • AI Advisor — What it may say, what it refuses to say, and why
  • Roadmap — The thirteen phases and the data blocker on each

Advanced Topics

  • Contract Testing — Self-describing, self-testing APIs with MCP
  • AI Assistant — KAgent AI agents embedded in Backstage
  • Agentic Development Platform (ADP) — Agent-driven dev workflow + ops, HiTL approval gate, opt-in via bootstrap-ai.sh --adp
  • Agent Approvals — Human-in-the-loop gate for agent-initiated mutating actions: policy, approval API, Backstage UI
  • DORA & FinOps — DORA entity tab (Elite/High/Medium/Low badges) + FinOps cost overview with team dimension in Backstage
  • Security — Pod Security Standards, OPA/Gatekeeper, RBAC, per-team secret isolation, production hardening
  • Flaky-Test Quarantine — How flaky tests get detected (exporter) and acted on (auto quarantine PRs)
  • Test-Impact Analysis — Selective test execution on PRs via pytest-testmon (python-service golden path)

Operations

Shift-Left Programme

Reference

  • Roadmap — Upcoming features and milestones, tracked as GitHub Project issues
  • Architecture — System design and data flow