Architecture runtime topology

Architecture Runtime Topology

Bundled commands use $PLUGIN_ROOT ($env:PLUGIN_ROOT in PowerShell; same path suffix) for the plugin root. Set it once: use the host's plugin-root variable when defined (Claude Code: PLUGIN_ROOT="$CLAUDE_PLUGIN_ROOT"), otherwise the absolute path of this plugin's root directory.

Use when the question is how the system runs, communicates, fails, scales, or is operated.

Inputs

• Deployment: Docker, Compose, Kubernetes, Helm, Terraform, serverless, Procfile, platform manifests, CI/CD, release config.
• Runtime config: env/config directories, service discovery, feature flags, schedulers/jobs.
• Integrations: HTTP, gRPC, GraphQL, queues, topics, cache, database, external APIs, webhooks, batch.
• Observability: logs, metrics, traces, dashboards, alerts, SLOs, health checks, incidents.
• Resilience: timeout, retry, circuit breaker, fallback, bulkhead, rate limit, idempotency, backpressure, degradation.
• Ownership: service owners, CODEOWNERS/OWNERS, runbooks, escalation paths, review gates.

Record paths and signal names only; never repeat secret values.

Probe

python3 "$PLUGIN_ROOT/scripts/architecture_probe.py" <repo-path> --json

Use runtime_topology as evidence inventory. It detects deployment artifacts and runtime signal names, not production truth.

Lenses

• Topology: what deploys, what calls what, and which runtime edges matter.
• Failure behavior: slow/down/partial/overloaded/inconsistent dependency response.
• Observability: critical paths have actionable logs, metrics, traces, health checks, owners.
• Deployment conformance: source boundaries and IaC/deployment boundaries agree or drift.
• Operability: teams can diagnose, roll back, and safely change runtime dependencies.
• Ownership: runtime-critical surfaces have source-backed owner or review path.
• Time-driven state: identify where state lives and which timer, refresh, lifecycle, scheduler, or storage signal re-evaluates it. Elapsed time alone is not an emission trigger. Use async-state-consistency for lifecycle semantics, replay/publication authority, linearization, and controlled race proof.
• Quality attributes: availability, reliability, latency, scalability, security, cost, operability.

Output

Compact: runtime summary, highest-impact coupling/resilience risks, evidence inspected, unknowns needing traces/config/owner input, next validation.

Durable: architecture_intelligence.runtime_topology.v1.

Do not infer production state from repo files alone. Do not treat an import as proof a tactic is correctly used. Recommend distributed-system patterns only when topology and quality-attribute scenarios justify them.