Cyber Resilience Framework: How to Build, Test & Improve It

Build a cyber resilience framework from MITRE CREF and NIST SP 800-160: strategy elements, a step-by-step plan, practices by objective, and how to test.

Blueprint of a cyber resilience framework over a construction site plan, with one section marked untested

Key Takeaways

  • MITRE CREF is MITRE’s cyber resilience framework. MITRE ATT&CK is a source of test scenarios. CREF shares its goals, objectives, and techniques with NIST SP 800-160 Vol. 2.

  • Business owners should set disruption tolerances in operational terms. Security should select the controls that support them.

  • NIST defines 14 cyber-resiliency techniques and says organizations cannot employ all of them simultaneously. Selection is therefore part of the design work.

  • A cyber resilience strategy remains credible only when tests show that people can Continue, Constrain, and Reconstitute under pressure.

A cyber resilience framework is the structured set of goals, objectives, practices, and test obligations an organization uses to keep critical services running through a cyberattack. It works as a loop: build it, test it, improve it. Skip the middle step and what you are holding is a hypothesis.

What Is a Cyber Resilience Framework?

A cyber resilience framework is the structure an organization uses to engineer and operate resilience. It names the goals the organization commits to, the objectives that make those goals concrete, the practices that deliver them, and the tests that show they hold. Its four underlying capabilities are anticipate, withstand, recover, and adapt. They are covered in Cyber Resilience: What It Is and Why Security Alone Isn’t Enough. This article starts where that definition stops.

The word "engineer" is deliberate. When MITRE released its Cyber Resiliency Engineering Framework Navigator in February 2023, Wen Masters, MITRE's vice president for cyber technologies, said that "resilience in the face of cyber threats must be engineered." A framework is how that engineering gets organized above the level of a single system.

It sits beside the organization’s security framework. A security framework asks whether controls exist and work. A resilience framework asks what keeps running when they fail, for how long, and who decides. Most organizations document their security framework more thoroughly than their resilience framework.

The Core Elements of a Cyber Resilience Strategy

A cyber resilience strategy decides which services must keep running, how much disruption each one can absorb, and what the organization will spend to hold that line. Seven building blocks turn that decision into a program somebody can operate:

  • Governance and ownership. One executive owns the program and its budget. Each critical service also gets a business owner with the authority to accept or reject a tolerance.

  • Critical-service mapping. List the services the business delivers, then the systems behind them. A general contractor delivers won bids and finished buildings. Its estimating platform matters because of the bid.

  • Tolerance for disruption. For each service, state the longest outage and the largest data loss the business will accept, in hours and in records. Agree on it while nobody is under pressure.

  • Dependency and third-party mapping. Trace each service to the people, vendors, and connections it needs. In partner-heavy sectors, most of that map is other companies.

  • Capability investments. Fund the technique that advances a specific objective for a specific service. Catalog coverage alone does not show that the required capability will work for that service.

  • Testing cadence. Decide how often each service’s claims will be tested and against which scenario class. Until a tolerance is tested, it is supported by documentation rather than observed evidence.

  • Improvement loop. Route every test finding to an owner, a date, and the building block it changes. A finding is not resolved until it changes part of the program.

Ownership is the block the other six depend on. Without a named business owner, security may set a tolerance that the organization revises during the first outage.

How to Build a Cyber Resilience Plan

A cyber resilience plan turns the strategy into named services, numbers, and owners for one organization. The sequence below follows a mid-size general contractor with an in-house engineering group and about a dozen active project sites. The firm is a composite, drawn from how these businesses run.

  1. Identify critical services. Three rise to the top: bid preparation and estimating; project-site operations, meaning site networks, document control, and BIM and design files; and payroll for site crews and subcontractors.

  2. Map dependencies. Estimating depends on a cloud estimating platform, the design-file server, and supplier price feeds. Site operations depend on temporary networks and on remote access for subcontractors. Payroll depends on an outside processor and on timesheets that supervisors submit from the field.

  3. Set tolerances. A bid due Friday at 14:00 cannot move, so in bid week estimating tolerates hours of downtime. Crews expect pay on Thursday. Design files can sit offline for a day, but silent corruption is unacceptable.

  4. Assign owners. The chief estimator owns estimating. Each project director owns a site. The finance director owns payroll. The CISO owns the method: framework, controls, and test schedule.

  5. Choose controls per CREF objective. For design files, Constrain and Reconstitute come first: segment site networks from the design-file server and keep protected backups. For payroll, Continue comes first: a manual fallback for one pay cycle.

  6. Write the plan. Give each service two pages: service, tolerance, dependencies, owner, prioritized objectives, and chosen controls. This plan stays at the program level. The incident playbook is covered separately in CISO Cyber Incident Response Plan: Prepare, Test & Respond.

  7. Schedule the first test. Put a date on the calendar before anyone signs the plan. Start with the service whose tolerance you trust least.

Step 7 is the one most plans leave out. Without a scheduled test, the plan documents intent but provides no evidence that it works.

Cyber Resilience Frameworks: NIST, MITRE ATT&CK and Other Approaches

One classification first. MITRE describes ATT&CK as "a globally-accessible knowledge base of adversary tactics and techniques based on real-world observations." That makes it a catalog of adversary behavior, and a very good one. It describes how attackers operate. It sets no resilience goals for the organization being attacked. MITRE's resilience framework is the Cyber Resiliency Engineering Framework (CREF), which MITRE says it originally created with NIST and which shares its constructs with NIST SP 800-160 Vol. 2.

Most cyber resilience strategies borrow from more than one reference framework, because each answers a different question:

Reference

Purpose

Level

What it provides

What it does not provide

NIST SP 800-160 Vol. 2 Rev. 1 (December 2021)

Engineering cyber-resilient systems

System, within mission and business context

Goals, objectives, 14 techniques, 50 implementation approaches, design principles

An operating model, named owners, or a test schedule

MITRE CREF and CREF Navigator (released February 2023)

A searchable, visual form of the same constructs

System and program

The working vocabulary: 4 goals, 8 objectives, selectable techniques, links to ATT&CK techniques and mitigations

Tolerances, governance, or evidence that any of it works

NIST CSF 2.0 (February 2024)

Cybersecurity outcomes taxonomy

Organization

Six functions (Govern, Identify, Protect, Detect, Respond, Recover) for mapping and board reporting

Engineering depth, or how to achieve any outcome

MITRE ATT&CK

Knowledge base of adversary tactics and techniques

Threat

Threat-informed scenarios grounded in observed behavior

Any resilience goal, objective, or program structure

NIST is explicit that selection is the design work: "Employment of all cyber resiliency techniques and approaches is not needed to achieve the cyber resiliency objectives prioritized by stakeholders. In fact, it is not possible to employ all techniques and approaches simultaneously."

When a program already reports in CSF terms, the mapping should run in one direction. CREF’s Recover goal aligns with CSF’s Recover function. Withstand extends across CSF’s Respond function and the Technology Infrastructure Resilience category under Protect (PR.IR). Incident Response Preparedness: How to Build & Test Readiness covers the operational preparation behind those outcomes. This is an editorial mapping, not a NIST crosswalk. The handling layer under Respond and Recover sits in the revised incident handling guidance. Organizations measured against the UK NCSC’s Cyber Assessment Framework have a separate exercise in Cyber Resilience Assessment: How to Measure Your Readiness.

The working stack for a mid-size firm: CREF for vocabulary, 800-160 when engineers need depth, CSF for the board, ATT&CK for scenarios.

Cyber Resilience Best Practices for Security Teams

The cyber resilience best practices that hold up are the ones tied to a named objective, because the objective defines what to test. CREF lists eight objectives. Seven drive day-to-day practice. The eighth, Re-Architect, is a design decision you revisit when a major platform changes. Concrete cyber resilience examples from one engineering firm:

  • Prevent or Avoid: scope subcontractor access to the job. NIST frames the objective as precluding "the successful execution of an attack." The contractor issues site-scoped, time-limited accounts that expire when a trade leaves the project, in place of standing VPN access for every subcontractor's office.

  • Prepare: keep a course of action per critical service. The objective asks for "a set of realistic courses of action." For bid week, that means a written fallback: which estimator works from which offline copy, and who may request a bid extension.

  • Continue: run payroll in degraded mode. The fallback is last cycle's register, a signed variance process, and a contact at the bank who knows the procedure exists.

  • Constrain: segment every site. Each temporary site network is cut off from the design-file server. A compromised site laptop reaches drawings for one project at most.

  • Reconstitute: restore in business order. Protected offline backups of BIM and design files come back in the order project directors set in advance: active bids, then live construction, then archive.

  • Understand: update the dependency map at every kickoff. Each new project adds subcontractors, site connections, and data flows that the previous map does not include.

  • Transform: change the process after each finding. When a test exposes a timesheet bottleneck, the firm changes how supervisors submit hours, in addition to fixing the system behind it.

Patching and MFA sit underneath every item on that list. The practice is the objective-specific decision above them. What happens after a test matters as much as the test, and there is a longer treatment of turning exercise findings into change.

The Missing Step: Testing Whether Your Cyber Resilience Framework Actually Works

A framework defines how resilience should work, but testing is needed to show whether its claims hold. Some claims can be checked on paper: a backup exists, a segment is configured, or an owner is named.

Others only exist when people carry them out. Continue, Constrain, and Reconstitute fall in that group. The payroll fallback works only if the finance director reaches for it at 06:00 on a Thursday. The site segment holds only if nobody opens a firewall rule to get a delayed project moving. The restore order holds only if three project directors keep to it while all three are calling the CISO.

NIST already puts testing inside the design. SP 800-160 Vol. 2 ties the Prepare objective to "contingency planning, continuity of operations plan (COOP), training, exercises, and incident response and recovery plans." So the test stage needs rules of its own: test the claims that depend on people, set the failure condition before the test starts, and begin with the weakest tolerance.

Methods differ in what they can prove, and Cyber Resilience Testing: How CISOs Can Prove Resilience compares the available testing methods. If the current exercise is a slide deck delivered once a year, start by fixing a stale tabletop program. EU financial entities also carry specific obligations under DORA resilience testing. This test stage is the part of the lifecycle Reflex Security is built for.

From Framework to Live Crisis Simulation With Reflex Security

The requirement is a scenario built from the organization’s context. A generic ransomware scenario cannot test organization-specific dependencies, decision paths, or business consequences. Reflex’s One-Click Scenario Generation takes a company domain and returns five customized scenarios in 10 to 15 minutes. It builds them from public information an attacker could collect, including job postings, DNS records, vendor relationships, executive profiles, and news coverage. For the contractor, that means a scenario reflecting its subcontractors and bid calendar.

The requirement is to test claims that depend on people under pressure. The Adaptive Adversary Engine reacts to the team’s decisions instead of following a script. If the finance director delays the payroll fallback, the scenario continues and records the consequences.

The requirement is evidence that the improvement stage can use. The Reflex After Action Report traces each finding to the relevant moment in the simulation and supports it with transcript evidence. It includes MITRE ATT&CK and NIST CSF alignment, along with action items carrying owners and deadlines. The report is designed to support testing evidence for SOC 2, ISO 27001, and cyber insurance. The organization can then relate each finding to the CREF objective it tested and use the report as an input to the next design cycle.

See how Reflex runs a simulation, then test the service whose tolerance you trust least.

See what your team does under real pressure.

Put your readiness to the test. Sixty minutes, your real environment, no slides.

Book A demo

Frequently Asked Questions

What is the difference between a cyber resilience framework and a cybersecurity framework?

A cybersecurity framework organizes cyber-risk outcomes across governance, protection, detection, response, and recovery. A cyber resilience framework focuses on sustaining and restoring critical services when adversity persists or controls fail. NIST CSF 2.0 is a cybersecurity framework. NIST SP 800-160 Vol. 2 and MITRE CREF provide cyber-resiliency engineering constructs.

Which cyber resilience framework should a mid-size company start with?

A midsize company can start with MITRE CREF for vocabulary and retain its existing NIST CSF 2.0 profile for reporting. The CREF Navigator is free and organizes the goals, objectives, and techniques on one screen. Bring in the full NIST SP 800-160 Vol. 2 when engineers begin designing systems around specific objectives. The publication itself says employing every technique it lists is neither needed nor possible.

Can MITRE ATT&CK be part of a cyber resilience program?

Yes, as an input. ATT&CK describes how real adversaries behave. That makes it a useful source for test scenarios. An organization can select techniques used against similar firms and test whether its Constrain and Reconstitute controls hold against them. MITRE has already linked the CREF Navigator to ATT&CK techniques and mitigations, so the mapping work is partly done.

Who should own the cyber resilience framework: the CISO or the business?

Both, for different parts. The business owns the tolerances, because only the business can say how long a bid platform or a payroll run can be down. The CISO owns the method: framework choice, controls, and test schedule. When the CISO sets tolerances alone, they tend to reflect what security can deliver rather than what the business can survive.

How often should a cyber resilience framework be reviewed?

Review the whole framework once a year. Review a single service's section sooner whenever something material changes: a major new subcontractor, an acquisition, a new site model, or a test that missed its tolerance. The cadence is an organizational decision, so write it into the plan where it survives staff turnover.

Does a cyber resilience framework need board approval?

It depends on the organization’s sector. DORA, for example, makes the management body of an EU financial entity responsible for defining and approving its ICT risk management framework. Where no rule applies, the board should still approve the tolerances because they are business risk decisions. Without that review, the tolerances may remain unchallenged until an outage exposes their operational consequences.

References

MITRE. (n.d.). CREF Navigator. MITRE. https://crefnavigator.mitre.org/

MITRE. (n.d.). MITRE ATT&CK. MITRE. https://attack.mitre.org/

MITRE. (2023, February 2). MITRE launches Cyber Resiliency Engineering Framework Navigator. MITRE. https://www.mitre.org/news-insights/news-release/mitre-launches-cyber-resiliency-engineering-framework-navigator

National Institute of Standards and Technology. (2024, February 26). The NIST cybersecurity framework (CSF) 2.0 (NIST CSWP 29). National Institute of Standards and Technology. https://nvlpubs.nist.gov/nistpubs/CSWP/NIST.CSWP.29.pdf

European Parliament & Council of the European Union. (2022, December 14). Regulation (EU) 2022/2554 on digital operational resilience for the financial sector. EUR-Lex. https://eur-lex.europa.eu/eli/reg/2022/2554/oj

Ross, R., Pillitteri, V., Graubart, R., Bodeau, D., & McQuaid, R. (2021, December 9). Developing cyber-resilient systems: A systems security engineering approach (NIST SP 800-160 Vol. 2 Rev. 1). National Institute of Standards and Technology. https://doi.org/10.6028/NIST.SP.800-160v2r1

Cite this article

Marlow Bryant.

Cyber Resilience Framework: How to Build, Test & Improve It.

Reflex Security.

https://reflexsecurity.com/guides/cyber-resilience-framework

Incident closed // after-action ready

Incident closed

See what your team does

under real pressure.

True cyber crisis readiness isn't built in a slide deck—it's forged in the trenches. Every team believes it’s ready. The simulation is where you find out. Sixty minutes, your real environment, no slides.

Book A demo

ReflexSecurity

AI-powered incident-response simulation platform.

AI-powered incident-response simulation platform.

© 2026 Reflex Security. All rights reserved.