The open source toolbox
for resilient operations

Rebound is a set of tools, ranging from a network injection fault proxy to a reliability automation platform, so people can rely on you being there when they need.

Get Started |
curl -sSL https://rebound.how/install/full.sh | bash

Copied!

What is Rebound?

Deliver products people can rely on

  • System Agnostic

    Rebound's toolbox is not bound to a particular system. Whether it's Cloud, On-Premise or "It's complicated", Rebound follows.

  • Metrics Driven

    With knowledge comes wisdom. Rebound has a mission to turn resilience from a perceived cost to a must-have strategic asset.

  • Scales With You

    Grow at your own pace. Rebound adapts to your journey—from early development to enterprise operations.

  • No Lock-In

    Naturally Open-Source. Invest safely in the future by owning the stack.

  • Engineers First

    Useful not painful. Rebound encourages resilience thinking at all stages without being obnoxious.

Products

Unfault

Follow your code breadcrumbs from the CLI

  • See How Functions Connect

    Unfault leaves context breadcrumbs across your codebase so you can orient yourself fast and stop paging code into your head.

  • Follow The Route

    Trace how logic flows through your app - from routes and entry points to the functions they reach.

  • Missing Guardrails

    Highlights resilience hardening gaps like timeouts, retries, and circuit breakers.

  • Silent Failure Paths

    Finds places where errors vanish without a trace - missing logging, tracing, or correlation IDs. 

$ unfault review

Looks good overall, with a couple spots that deserve a closer look. Two themes
keep showing up: observability (logging and tracing) and resilience hardening.
Starting point: reliably_app/login/service.py (Unstructured logging:
logger.warning); then reliably_app/notification/tasks (HTTP call to external
service in `publish_notification` lacks circu...). If an incident hits,
correlation IDs and structured logs make the follow-up a lot calmer.

At a glance
  · One call missing a timeout
  · Circuit breakers would help fail fast when dependencies are down

────────────────────────────────────────────────────────────────────────────────
6062ms - python / fastapi - 185 files
Tip: use --output full to drill into hotspots.
Read the docs

fault

Developer-Friendly Reliability

  • Lightning-Fast

    A Rust-powered CLI that simulates network disruptions right on your local machine.

  • Real-World Scenarios

    Easily configure tests via YAML to mimic latencies, bandwidth constraints, packet loss, and HTTP errors.

  • CI Integration

    Automatically run these tests within your pipeline and receive detailed, actionable reports. 

$ fault scenario run --scenario scenario.yaml

================ Running Scenarios ================

⡏  4/4  Latency Increase By 30ms Steps From Downstream  ▮▮▮▮
⡏  1/1  Within Allowed Latency While Bandwidth At 5 bytes/second  ▮
⡏  1/1  Circuit Breaker Takes Care of 404  ▮
⡏  1/1  Packet loss has no impact on service performance  ▮

===================== Summary =====================

Tests run: 7, Tests failed: 2
Total time: 2.4s

Report saved as report.json
$ fault scenario run --scenario scenario.yaml

================ Running Scenarios ================

⡏  4/4  Latency Increase By 30ms Steps From Downstream  ▮▮▮▮
⡏  1/1  Within Allowed Latency While Bandwidth At 5 bytes/second  ▮
⡏  1/1  Circuit Breaker Takes Care of 404  ▮
⡏  1/1  Packet loss has no impact on service performance  ▮

===================== Summary =====================

Tests run: 7, Tests failed: 2
Total time: 2.4s

Report saved as report.json
Read the docs

Chaos Toolkit

Reliability Policies in Action

  • Executable Policies

    Convert your resilience objectives into clear, measurable tests.

  • Clear Feedback

    Gain actionable insights that let you address vulnerabilities before they escalate.

  • Fully Customizable

    Cover all your systems, whether it's legacy, Cloud, On-Premise or "It's complicated", Rebound follows you. 

$ chaos run policy.json
[2025-03-04 15:23:27 INFO] Validating the experiment's syntax
[2025-03-04 15:23:28 INFO] Experiment looks valid
[2025-03-04 15:23:28 INFO] Running experiment: Loss of pod capacity does not impact latency
[2025-03-04 15:23:28 INFO] Steady-state strategy: after-method-only
[2025-03-04 15:23:28 INFO] Rollbacks strategy: default 
[2025-03-04 15:23:28 INFO] Playing your experiment's method now...
[2025-03-04 15:23:28 INFO] Action: Run 5qps to home page for 60s [in background]
[2025-03-04 15:23:28 INFO] Action: Delete application pod
[2025-03-04 15:23:28 INFO] Action: Let it self-heal
[2025-03-04 15:23:28 INFO] Pausing activity for 45s or until the execution is resumed
[2025-03-04 15:24:13 INFO] Resuming execution...
[2025-03-04 15:24:28 INFO] Steady state hypothesis: Steady-State Hypothesis
[2025-03-04 15:24:28 INFO] Probe: Capacity is back to normal
[2025-03-04 15:24:28 INFO] Probe: Latency was not impacted
[2025-03-04 15:24:28 CRITICAL] Steady state probe 'Latency was not impacted' is not in the given tolerance
[2025-03-04 15:24:28 INFO] Let's rollback...
[2025-03-04 15:24:28 INFO] No declared rollbacks, let's move on.
[2025-03-04 15:24:28 INFO] Experiment ended with status: deviated
[2025-03-04 15:24:28 INFO] The steady-state has deviated, a weakness may have been discovered
$ chaos run policy.json
[2025-03-04 15:23:27 INFO] Validating the experiment's syntax
[2025-03-04 15:23:28 INFO] Experiment looks valid
[2025-03-04 15:23:28 INFO] Running experiment: Loss of pod capacity does not impact latency
[2025-03-04 15:23:28 INFO] Steady-state strategy: after-method-only
[2025-03-04 15:23:28 INFO] Rollbacks strategy: default 
[2025-03-04 15:23:28 INFO] Playing your experiment's method now...
[2025-03-04 15:23:28 INFO] Action: Run 5qps to home page for 60s [in background]
[2025-03-04 15:23:28 INFO] Action: Delete application pod
[2025-03-04 15:23:28 INFO] Action: Let it self-heal
[2025-03-04 15:23:28 INFO] Pausing activity for 45s or until the execution is resumed
[2025-03-04 15:24:13 INFO] Resuming execution...
[2025-03-04 15:24:28 INFO] Steady state hypothesis: Steady-State Hypothesis
[2025-03-04 15:24:28 INFO] Probe: Capacity is back to normal
[2025-03-04 15:24:28 INFO] Probe: Latency was not impacted
[2025-03-04 15:24:28 CRITICAL] Steady state probe 'Latency was not impacted' is not in the given tolerance
[2025-03-04 15:24:28 INFO] Let's rollback...
[2025-03-04 15:24:28 INFO] No declared rollbacks, let's move on.
[2025-03-04 15:24:28 INFO] Experiment ended with status: deviated
[2025-03-04 15:24:28 INFO] The steady-state has deviated, a weakness may have been discovered
Read the docs

Reliably

Orchestrate Your Resilience Strategy

  • Centralized Management

    Coordinate and monitor resilience tests across your entire infrastructure from a single platform.

  • Unified Visibility

    Automatically map system health, track key resilience metrics, and identify improvement areas.

  • AI-Powered Policies

    Leverage GenAI to generate targeted tests and policies that evolve as your system does.

A screenshot from the Reliably application. It shows an expriment with a summary of the last execution results and a breakdown of the different actions the experiment is made of.
Read the docs
Why Rebound?

Features

    Context Breadcrumbs

    Stop paging code into your head - Unfault leaves runtime context breadcrumbs across your codebase.

    Hello

    See How Logic Flows

    Trace how functions connect and where code paths lead, without leaving your editor.

    Hello

    Ask Once

    Use graph traversal (not grep) to answer questions like “what’s the authentication flow?” and get the path.

    Hello

    Runtime Guardrails Review

    Spot missing timeouts, tracing/logging gaps, and resilience hardening opportunities before incidents.

    Hello

    Network Fault Injector

    Gain understanding of how your services react to poor network conditions.

    Hello

    Rapid Testing Scenarios

    Turn Reliability into non-regression scenarios from your CI build.

    Hello

    Reliability As Code

    Declare reliability tests as code stored like any other assets for review and visibility.

    Hello

    Comprehensive Systems Coverage

    Ensure all parts of your systems can be verified and kept in check, including your legacy systems.

    Hello

    Extensible By Design

    If something is amiss, easily extend for your own unique requirements.

    Hello

    Safety Included

    Remain in control and rapidly terminate tests when your system is under fire.

    Hello

    Teams Communication

    Let teams know when a reliability test is underway, via Slack, email or a webhook.

    Hello

    Reporting

    Build reports that can be easily shared throughout your organization.

    Hello

    Orchestration

    Orchestrate the fleet of reliability tests and policies across your system from a unified platform.

    Hello

    Periodic Scheduling

    Run tests once or periodically automatically.

    Hello

    Open Telemetry Support

    Make your tests fully part of your operations via Open Telemetry traces and metrics.

    Hello

    GenAI Assistant

    Use the gentle power of GenAI to help you create reliability tests.

    Hello

    Effort Scoring

    Keep an eye on engineering effort with a simple scoring system.

    Hello

    Template

    Create tests and policies template to let teams parameterized them in their own context.

    Hello

    Runs anywhere

    Rebound products can run on any system and are simple to deploy locally, as containers, in a Kubernets cluster or from a serverless ecosystem.

    Hello
Get started