Observability as code, continuously

Make your repo ready for production-quality SLAs - effortlessly.

HoundOS installs as a GitHub App, derives SLOs and SLIs from Observability Engineering best practices for your services, and opens pull requests that close the OpenTelemetry instrumentation gaps — then keeps them aligned as your code changes.

v0 preview · Java, Node, Python, Go · No telemetry leaves your repos

microservices-orderscommit a7c91e3

# Recommended SLOs

Order Entry Success Rate99.95% / 30d

Order Entry P95 Latency< 500ms / 24h

Kafka Produce Success99.9% / 1h

Consumer Lag< 1000 msgs / 5m

# Pull requests

#142 OTel SDK + Kafka auto-instropen

#143 Custom validation spansopen

#144 SLO contractmerged

How it works

Install on your repos

HoundOS reads the codebase, detects services and runtimes, and maps message brokers, databases, and business flows.

Get SLOs grounded in Observability Engineering best practices

A curated template library produces a defensible starting set of SLOs and SLIs per service archetype. Edit, version, and approve as the contract.

Ship instrumentation as PRs

HoundOS opens pull requests that add OpenTelemetry auto-instrumentation and custom spans/metrics to close the gaps.

Continuous adherence

Every push to the default branch re-runs analysis. If a change breaks the instrumentation backing an agreed SLI, HoundOS fails the PR check and surfaces a drift item with the one-line fix.

Drift detected on PR #267

ValidationsAggregatorService.handle() lost the order.validation.passed attribute. SLI Order Entry Success Rate can no longer be computed from traces.

Suggested fixRe-add the attribute or accept the SLO downgrade.