Steadybit
Chaos engineering as governed tests, with the first MCP server for the discipline.
About
Steadybit is a chaos engineering and reliability management platform founded in 2019 in Solingen, Germany. Its angle is governance: instead of ad hoc fault scripts, you model a hypothesis, constrain blast radius with environments and teams, run attacks against discovered targets, and use checks to decide whether a run passes, fails, or aborts, so chaos experiments behave like CI quality gates with evidence beside your other test results. The agent auto-detects reliability risks and suggests fixes before you run anything.
The company raised a $7.8M seed led by Boldstart Ventures in 2022 and a $6M Series A in March 2024 (Boldstart, Paladin Capital Group, Angular Ventures). In 2025 it launched what it calls the first MCP server for chaos engineering, letting teams pull experiment insights, gap analyses, and experiment suggestions into LLM workflows; Salesforce's engineering team publicly described using it with PagerDuty and Datadog data. The platform core is SaaS with open-source extensions for integrations.
What it can do
- Hypothesis-driven chaos experiments with blast-radius controls
- Reliability Hub with prebuilt attacks and advice
- Automatic reliability risk detection with recommended fixes
- Open-source extensions for clouds, observability, load testing, meshes
- MCP server for chaos engineering in LLM workflows
- GitHub Action and API/CLI for CI integration
- Scheduling and steady-state checks
- Professional and Enterprise plans with 30-day trial
Setup
- Sign up at steadybit.com and deploy the Steadybit agent into your environments.
- Review the auto-detected reliability risks and recommended fixes.
- Build experiments in the no-code editor with hypothesis, targets, and blast-radius constraints.
- Run manually, on a schedule, or in CI via the GitHub Action; optionally connect the MCP server to your LLM workflows.
Prerequisites
Environments where you can deploy the Steadybit agent (cloud, on-premise, or Kubernetes).
Use cases
Works with
Trust score — 7/10
Scored 7/10 (TIER metrics, metrics-v1). Evidence: founded 2019 (Germany); ~$15M raised ($7.8M seed + $6M Series A, Boldstart-led); first MCP server for chaos engineering (2025) with Salesforce engineer on record; open-source extension ecosystem; active platform in 2026. Strengths: governed-test framing fits CI culture, genuine MCP/LLM integration, extensibility. Docked for: (1) smaller company (11-50 people) in a category led by Gremlin; (2) no permanent free tier, pricing gated. License: proprietary platform, open-source extensions.
How the trust score is built
Full methodology →- Adoption · stars/downloads, log-scaled vs peers
- 6.0/10
- Maintenance · recency, releases, issue load
- 7.0/10
- Openness · license + docs
- 8.0/10
- Manual review · sanity read + red-flag check
- 7.0/10
- · Founded 2019, Solingen Germany
- · ~$15M raised: $7.8M seed (2022, Boldstart) + $6M Series A (Mar 2024)
- · First MCP server for chaos engineering (2025); Salesforce engineer on record using it
- · Open-source extension ecosystem; governed-test/CI-quality-gate framing
- · Small company (11-50) vs Gremlin; no permanent free tier
Sources
Every fact above was checked against these sources. Where a claim comes from the provider and couldn't be independently verified, the text says so.