schema-details.json

{
  "activities": {
    "SM.1.A": {
      "description": "Interview business owners and stakeholders and create a list of worst-case scenarios across the organization’s various application and data assets.  Based on the way in which your organization builds, uses, or sells software, the list of worst-case scenarios can vary widely, but common issues include data theft or corruption, service outages, monetary loss, reverse engineering, account compromise, etc. After broadly capturing worst-case scenario ideas, collate and select the most important based on collected information and knowledge about the core business.  Any number can be selected, but aim for at least 3 and no more than 7 to make efficient use of time and keep the exercise focused. Elaborate a description of each of the selected items and document details of contributing worst-case scenarios, potential contributing factors, and potential mitigating factors for the organization. The final business risk profile should be reviewed with business owners and other stakeholders for understanding.",
      "objective"  : "Establish unifed strategic roadmap for software security within the organization",
      "proof"      : "Do the business stakeholders understand your organization’s risk profile?"
    },
    "SM.1.B": {
      "description": "Understanding the main business risks to the organization, evaluate the current performance of the organization against each the twelve Practices. Assign a score for each Practice from 1, 2, or 3 based on the corresponding Objective if the organization passes all the cumulative success metrics. If no success metrics are being met, assign a score of 0 to the Practice. Once a good understanding of current status is obtained, the next goal is to identify the Practices that will be improved in the next iteration. Select them based on business risk profile, other business drivers, compliance requirements, budget tolerance, etc. Once Practices are selected, the goals of the iteration are to achieve the next Objective under each. Iterations of improvement on the assurance program should be approximately 3-6 months, but an assurance strategy session should take place at least every 3 months to review progress on activities, performance against success metrics and other business drivers that may require program changes.",
      "objective"  : "Establish unified strategic roadmap for software security within the organization",
      "proof"      : "Is there a software security assurance program in place? Are development staff aware of future plans for the assurance program?"
    },
    "SM.2.A": {
      "description": "Establish a simple classification system to represent risk-tiers for applications.  In its simplest form, this can be a High/Medium/Low categorization.  More sophisticated classifications can be used, but there should be no more than seven categories and they should roughly represent a gradient from high to low impact against business risks.   Working from the organization’s business risk profile, create project evaluation criteria that maps each project to one of the risk categories.  A similar but separate classification scheme should be created for data assets and each item should be weighted and categorized based on potential impact to business risks. Evaluate collected information about each application and assign each a risk category based upon overall evaluation criteria and the risk categories of data assets in use.  This can be done centrally by a security group or by individual project teams through a customized questionnaire to gather the requisite information. An ongoing process for application and data asset risk categorization should be established to assign categories to new assets and keep the existing information updated at least biannually.",
      "objective"  : "Measure relative value of data and software assets and choose risk tolerance",
      "proof"      : "Are many of your applications and resources categorized by risk?"
    },
    "SM.2.B": {
      "description": "With a classification scheme for the organization’s application portfolio in place, direct security goals and assurance program roadmap choices can be made more granular. The assurance program’s roadmap should be modified to account for each application risk category by specifying emphasis on particular Practices for each category.  For each iteration of the assurance program, this would typically take the form of prioritizing more higher-level Objectives on the highest risk application tier and progressively less stringent Objectives for lower/other categories. This process establishes the organization’s risk tolerance since active decisions must be made as to what specific Objectives are expected of applications in each risk category.  By choosing to keep lower risk applications at lower levels of performance with respect to the Security Practices, resources are saved in exchange for acceptance of a weighted risk.  However, it is not necessary to arbitrarily build a separate roadmap for each risk category since that can leads to inefficiency in management of the assurance program itself.",
      "objective"  : "Measure relative value of data and software assets and choose risk tolerance",
      "proof"      : "Are risk ratings used to tailor the required assurance activities?"
    },
    "SM.3.A": {
      "description": "Research and gather information about security costs from intra-industry communication forums, business analyst and consulting firms, or other external sources.  In particular, there are a few key factors that need to be identified. First, use collected information to identify the average amount of security effort being applied by similar types of organizations in your industry.  This can be done either top-down from estimates of total percentage of budget, revenue, etc. or it can be done bottom-up by identifying security-related activities that are considered normal for your type of organization.  Overall, this can be hard to gauge for certain industries, so collect information from as many relevant sources as are accessible. The next goal of researching security costs is to determine if there are potential cost savings on third-party security products and services that your organization currently uses. When weighing the decision of switching vendors, account for hidden costs such as retraining staff or other program overhead. Overall, these cost-comparison exercises should be conducted at least annually prior to the subsequent assurance program strategy session. Comparison information should be presented to stakeholders in order to better align the assurance program with the business.",
      "objective"  : "Align security expenditure with relevant business indicators and asset value",
      "proof"      : "Is per-project data for the cost of assurance activities collected?"
    },
    "SM.3.B": {
      "description": "Collect project-specific information on the cost of past security incidents. For instance, time and money spent in cleaning up a breach, monetary loss from system outages, fines and fees to regulatory agencies, project-specific one-off security expenditures for tools or services, etc. Using the application risk categories and the respective prescribed assurance program roadmaps for each, a baseline security cost for each application can be initially estimated from the costs associated with the corresponding risk category. Combine the application-specific cost information with the general cost model based on risk category, and then evaluate projects for outliers, i.e. sums disproportionate to the risk rating.  These indicate either an error in risk evaluation/classification or the necessity to tune the organization’s assurance program to address root causes for security cost more effectively. The tracking of security spend per project should be done quarterly at the assurance program strategy session, and the information should be reviewed and evaluated by stakeholders at least annually.  Outliers and other unforeseen costs should be discussed for potential affect on assurance program roadmap.",
      "objective"  : "Align security expenditure with relevant business indicators and asset value",
      "proof"      : "Does your organization regularly compare your security spend with that of other organizations?"
    },
    "PC.1.A": {
      "description": "While an organization might have a wide variety of compliance requirements, this activity is specifically oriented around those that either directly or indirectly affect the way in which the organization builds or uses software and/or data. Leverage internal staff focused on compliance if available. Based on the organization’s core business, conduct research and identify third-party regulatory standards with which compliance is required or considered an industry norm.  Possibilities include the Sarbanes-Oxley Act (SOX), the Payment Card Industry Data Security Standards (PCI-DSS), the Health Insurance Portability and Accountability Act (HIPAA), etc. After reading and understanding each third-party standard, collect specific requirements related to software and data and build a consolidated list that maps each driver (third-party standard) to each of its specific requirements for security. At this stage, try to limit the amount of requirements by dropping anything considered optional or only recommended. At a minimum, conduct research at least biannually to ensure the organization is keeping updated on changes to third-party standards. Depending upon the industry and the importance of compliance, this activity can vary in effort and personnel involvement, but should always be done explicitly.",
      "objective"  : "Understand relevant governance and compliance drivers to the organization",
      "proof"      : "Do project stakeholders know their project’s compliance status?"
    },
    "PC.1.B": {
      "description": "Based upon the consolidated list of software and data-related requirements from compliance drivers, elaborate the list by creating a corresponding response statement to each requirement.  Sometimes called control statements, each response should capture the concept of what the organization does to ensure the requirement is met (or to note why it does not apply). Since typical audit practice often involves checking a control statement for sufficiency and then measuring the organization against the control statement itself, it is critical that they accurately represent actual organizational practices. Also, many requirements can be met by instituting simple, lightweight process elements to cover base-line compliance prior to evolving the organization for better assurance down the road. Working from the consolidated list, identify major gaps to feed the future planning efforts with regard to building the assurance program. Communicate information about compliance gaps with stakeholders to ensure awareness of the risk from non-compliance. At a minimum, update and review control statements with stakeholders at least biannually.  Depending on the number of compliance drivers, it may make sense to perform updates more often.",
      "objective"  : "Understand relevant governance and compliance drivers to the organization",
      "proof"      : "Are compliance requirements specifically considered by project teams?"
    },
    "PC.2.A": {
      "description": "Beginning with a current compliance guidelines, review regulatory standards and note any optional or recommended security requirements.  Also, the organization should conduct a small amount of research to discover any potential future changes in compliance requirements that are relevant. Augment the list with any additional requirements based on known business drivers for security.  Often it is simplest to consult existing guidance being provided to development staff and gather a set of best practices. Group common/similar requirements and rewrite each group as more generalized/simplified statements that meet all the compliance drivers as well as provide some additional security value.  Work through this process for each grouping with the goal of building a set of internal policies and standards that can be directly mapped back to compliance drivers and best practices. It is important for the set of policies and standards to not contain requirements that are too difficult or excessively costly for project teams to comply.  A useful heuristic is that approximately 80% of projects should be able to comply with minimal disruption.  This requires a good communications program being set up to advertise the new policies/standards and assist teams with compliance if needed.",
      "objective"  : "Establish security and compliance baseline and understand per-project risks",
      "proof"      : "Does the organization utilize a set of policies and standards to control software development?"
    },
    "PC.2.B": {
      "description": "Create a simple audit process for project teams to request and receive an audit against internal standards.  Audits are typically performed by security auditors but can also be conducted by security-savvy staff as long as they are knowledgeable about the internal standards. Based upon any known business risk indicators, projects can be prioritized concurrently with audit queue triage such that high-risk software is assessed sooner or more frequently.  Additionally, low-risk projects can have internal audit requirements loosened to make the audit practice more cost-effective. Overall, each active project should undergo an audit at least biannually.  Generally, subsequent audits after the initial will be simpler to perform if sufficient audit information about the application is retained. Advertise this service to business owners and other stakeholders so that they may request an audit for their projects.  Detailed pass/fail results per requirement from the internal standards should be delivered to project stakeholders for evaluation.  Where practical, audit results should also contain explanations of impact and remediation recommendations.",
      "objective"  : "Establish security and compliance baseline and understand per-project risks",
      "proof"      : "Are project teams able to request an audit for compliance with policies and standards?"
    },
    "PC.3.A": {
      "description": "Once an organization has established internal standards for security, the next level of enforcement is to set particular points in the project life-cycle where a project cannot pass until it is audited against the internal standards and found to be in compliance. Usually, the compliance gate is placed at the point of software release such that they are not allowed to publish a release until the compliance check is passed. It is important to provide enough time for the audit to take place and remediation to occur, so generally the audit should begin earlier, for instance when a release is given to QA. Despite being a firm compliance gate, legacy or other specialized projects may not be able to comply, so an exception approval process must also be created.  No more than about 20% of all projects should have exception approval.",
      "objective"  : "Require compliance and measure projects against organization-wide policies and standards",
      "proof"      : "Are projects periodically audited to ensure a baseline of compliance with policies and standards?"
    },
    "PC.3.B": {
      "description": "Organizations conducting regular audits of project teams generate a large amount of audit data over time.  Automation should be utilized to assist in automated collection, manage collation for storage and retrieval, and to limit individual access to sensitive audit data. For many concrete requirements from the internal standards, existing tools such as code analyzers, application penetration testing tools, monitoring software, etc. can be customized and leveraged to automate compliance checks against internal standards.  The purpose of automating compliance checks is to both improve efficiency of audit as well as enable more staff to self-check for compliance before a formal audit takes place.  Additionally, automated checks are less error-prone and allow for lower latency on discovery of problems. Information storage features should allow centralized access to current and historic audit data per project.  Automation solutions must also provide detailed access control features to limit access to approved individuals with valid business purpose for accessing the audit data. All instructions and procedures related to accessing compliance data as well as requesting access privileges should be advertised to project teams.  Additional time may be initially required from security auditors to bootstrap project teams.",
      "objective"  : "Require compliance and measure projects against organization-wide policies and standards",
      "proof"      : "Does the organization systematically use audits to collect and control compliance evidence?"
    },
    "EG.1.A": {
      "description": "Either internally or externally sourced, conduct security training for technical staff that covers the basic tenets of application security. Generally, this can be accomplished via instructor-led training in 1-2 days or via computer-based training with modules taking about the same amount of time per developer. Course content should cover both conceptual and technical information. Appropriate topics include high-level best practices surrounding input validation, output encoding, error handling, logging, authentication, authorization. Additional coverage of commonplace software vulnerabilities is also desirable such as a Top 10 list appropriate to the software being developed (web applications, embedded devices, client-server applications, back-end transaction systems, etc.). Wherever possible, use code samples and lab exercises in the specific programming language(s) that applies. To rollout such training, it is recommended to mandate annual security training and then hold courses (either instructor-led or computer-based) as often as required based on development head-count.",
      "objective"  : "Offer development staff access to resources around the topics of secure programming and deployment",
      "proof"      : "Have developers been given high-level security awareness training?"
    },
    "EG.1.B": {
      "description": "For development staff, assemble a list of approved documents, web pages, and technical notes that provide technology-specific security advice. These references can be assembled from many publicly available resources on the Internet. In cases where very specialized or proprietary technologies permeate the development environment, utilize senior, security-savvy staff to build security notes over time to create such a knowledge base in an ad hoc fashion. Ensure management is aware of the resources and briefs oncoming staff about their expected usage. Try to keep the guidelines lightweight and up-to-date to avoid clutter and irrelevance. Once a comfort-level has been established, they can be used as a qualitative checklist to ensure that the guidelines have been read, understood, and followed in the development process.",
      "objective"  : "Offer development staff access to resources around the topics of secure programming and deployment",
      "proof"      : "Does each project team understand where to find secure development best-practices and guidance?"
    },
    "EG.2.A": {
      "description": "Conduct security training for staff that highlights application security in the context of each role’s job function.  Generally, this can be accomplished via instructor-led training in 1-2 days or via computer-based training with modules taking about the same amount of time per person. For managers and requirements specifiers, course content should feature security requirements planning, vulnerability and incident management, threat modeling, and misuse/abuse case design. Tester and auditor training should focus on training staff to understand and more effectively analyze software for security-relevant issues.  As such, it should feature techniques for code review, architecture and design analysis, runtime analysis, and effective security test planning. Expand technical training targeting developers and architects to include other relevant topics such as security design patterns, tool-specific training, threat modeling and software assessment techniques.  To rollout such training, it is recommended to mandate annual security awareness training and periodic specialized topics training. Course should be available (either instructor-led or computer-based) as often as required based on head-count per role.",
      "objective"  : "Educate all personnel in the software life-cycle with role-specific guidance on secure development",
      "proof"      : "Are those involved in the development process given role-specific security training and guidance?"
    },
    "EG.2.B": {
      "description": "Using either internal or external experts, make security-savvy staff available to project teams for consultation.  Further, this coaching resource should be advertised internally to ensure that staff are aware of its availability. The coaching staff can be created by recruiting experienced individuals within the organization to spend some percentage of their time, around 10% maximum, performing coaching activities. The coaches should communicate between one another to ensure they are aware of each other’s area of expertise and route questions accordingly for efficiency. While coaches can be used at any point in the software life-cycle, appropriate times to use the coaches include during initial product conception, before completion of functional or detailed design specification(s), when issues arise during development, test planning, and when operational security incidents occur.  Over time, the internal network of coaching resources can be used as points-of-contact for communicating security-relevant information throughout the organization as well as being local resources that have greater familiarity with the ongoing project teams than a purely centralized security team might.",
      "objective"  : "Educate all personnel in the software life-cycle with role-specific guidance on secure development",
      "proof"      : "Are stakeholders able to pull in security coaches for use on projects?"
    },
    "EG.3.A": {
      "description": "Building upon written resources on topics relevant to application security, create and advertise a centralized repository (usually an internal web site).  The guidelines themselves can be created in any way that makes sense for the organization, but an approval board and straightforward change control processes must be established. Beyond static content in the form of best-practices lists, tool-specific guides, FAQs, and other articles, the support portal should feature interactive components such as mailing lists, web-based forums, or wikis to allow internal resources to cross-communicate security relevant topics and have the information cataloged for future reference. The content should be cataloged and easily searchable based upon several common factors such as platform, programming language, pertinence to specific third party libraries or frameworks, life-cycle stage, etc.  Project teams creating software should align themselves early in product development to the specific guidelines that they will follow.  In product assessments, the list of applicable guidelines and product-related discussions should be used as audit criteria.",
      "objective"  : "Mandate comprehensive security training and certify personnel for baseline knowledge",
      "proof"      : "Is security-related guidance centrally controlled and consistently distributed throughout the organization?"
    },
    "EG.3.B": {
      "description": "Either per role or per training class/module, create and administer aptitude exams that test people for comprehension and utilization of security knowledge.  Typically, exams should be created based on the role-based curricula and target a minimum passing score around 75% correct.  While staff should be required to take applicable training or refresher courses annually, certification exams should be required biannually at a minimum. Based upon pass/fail criteria or exceptional performance, staff should be ranked into tiers such that other security-related activities could require individuals of a particular certification level to sign-off before the activity is complete, e.g. an uncertified developer cannot pass a design into implementation without explicit approval from a certified architect.  This provides granular visibility on an per-project basis for tracking security decisions with individual accountability.  Overall, this provides a foundation for rewarding or penalizing staff for making good business decisions regarding application security.",
      "objective"  : "Mandate comprehensive security training and certify personnel for baseline knowledge",
      "proof"      : "Are developers tested to ensure a baseline skill-set for secure development practices?"
    },
    "TA.1.A": {
      "description": "Based purely on the business purpose of each software project and the business risk profile (if available) identify likely worst-case scenarios for the software under development in each project team. This can be conducted using simple attack trees or through a more formal threat modeling process such as Microsoft’s STRIDE, Trike, etc. To build attack trees, identify each worst-case scenario in one sentence and label these as the high-level goals of an attacker. From each attacker goal identified, identify preconditions that must hold in order for each goal to be realized. This information should be captured in branches underneath each goal where each branch is either a logical AND or a logical OR of the statements contained underneath. An AND branch indicates that each directly attached child nodes must be true in order to realize the parent node. An OR branch indicates that any one of the directly attached child nodes must be true in order to achieve the parent node. Regardless of the threat modeling approach, review each current and historic functional requirement to augment the attack tree to indicate security failures relevant to each. Brainstorm by iteratively dissecting each failure scenario into all the possible ways in which an attacker might be able to reach one of the goals. After initial creation, the threat model for an application should be updated when significant changes to the software are made. This assessment should be conducted with senior developers and architects as well as one or more security auditors.",
      "objective"  : "Identify and understand high-level threats to the organization and individual projects",
      "proof"      : "Do projects in your organization consider and document likely threats?"
    },
    "TA.1.B": {
      "description": "Initially, conduct an assessment to identify all likely threats to the organization based on software projects. For this assessment, consider threats to be limited to agents of malicious intent and omit other risks such as known vulnerabilities, potential weaknesses, etc. Begin by generally considering external agents and their corresponding motivations for attack. To this list, add internal roles that could cause damage and their motivations for insider attack. Based on the architecture of the software project(s) under consideration, it can be more efficient to conduct this analysis once per architecture type instead of for each project individually since applications of architecture and business purpose will generally be susceptible to similar threats. This assessment should be conducted with business owners and other stakeholders but also include one or more security auditors for additional perspective on threats. In the end, the goal is to have a concise list of threat agents and their corresponding motivations for attack.",
      "objective"  : "Identify and understand high-level threats to the organization and individual projects",
      "proof"      : "Does your organization understand and document the types of attackers it faces?"
    },
    "TA.2.A": {
      "description": "Further considering the threats to the organization, conduct a more formal analysis to determine potential misuse or abuse of functionality. Typically, this process begins with identification of normal usage scenarios, e.g. use-case diagrams if available. If a formal abuse-case technique isn’t used, generate a set of abuse-cases for each scenario by starting with a statement of normal usage and brainstorming ways in which the statement might be negated, in whole or in part. The simplest way to get started is to insert the word “no” or “not” into the usage statement in as many ways as possible, typically around nouns and verbs. Each usage scenario should generate several possible abuse-case statements. Further elaborate the abuse-case statements to include any application-specific concerns based on the business function of the software. The ultimate goal is for the completed set of abuse statements to form a model for usage patterns that should be disallowed by the software. If desired, these abuse cases can be combined with existing threat models. After initial creation, abuse-case models should be updated for active projects during the design phase. For existing projects, new requirements should be analyzed for potential abuse, and existing projects should opportunistically build abuse-cases for established functionality where practical.",
      "objective"  : "Increase accuracy of threat assessment and improve granularity of per-project understanding",
      "proof"      : "Do project teams regularly analyze functional requirements for likely abuses?"
    },
    "TA.2.B": {
      "description": "Based on the established attacker profiles, identify a rating system to allow relative comparison between the threats. Initially, this can be a simple high-medium-low rating based upon business risk, but any scale can be used provided that there are no more than 5 categories. After identification of a rating system, build evaluation criteria that allow each threat to be assigned a rating. In order to do this properly, additional factors about each threat must be considered beyond motivation. Important factors include capital and human resources, inherent access privilege, technical ability, relevant goals on the threat model(s), likelihood of successful attack, etc. After assigning each threat to a rating, use this information to prioritize risk mitigation activities within the development life-cycle. Once built for a project team, it should be updated during design of new features or refactoring efforts.",
      "objective"  : "Increase accuracy of threat assessment and improve granularity of per-project understanding",
      "proof"      : "Do project teams use a method of rating threats for relative comparison?"
    },
    "TA.3.A": {
      "description": "Conduct an assessment of your software code-base and identify any components that are of external origin. Typically, these will include open-source projects, purchased COTS software, and online services which your software uses. For each identified component, elaborate attacker profiles for the software project based upon potential compromise of third-party components. Based upon the newly identified attacker profiles, update software threat models to incorporate any likely risks based upon new attacker goals or capabilities. In addition to threat scenarios, also consider ways in which vulnerabilities or design flaws in the third-party software might affect your code and design. Elaborate your threat models accordingly with the potential risks from vulnerabilities and knowledge of the updated attacker profile. After initially conducted for a project, this must be updated and reviewed during the design phase or every development cycle. This activity should be conducted by a security auditor with relevant technical and business stakeholders.",
      "objective"  : "Concretely tie compensating controls to each threat against internal and third-party software",
      "proof"      : "Do project teams specifically consider risk from external software?"
    },
    "TA.3.B": {
      "description": "Conduct an assessment to formally identify factors that directly prevent preconditions for compromise represented by the threat models. These mitigating factors are the compensating controls that formally address the direct risks from software. Factors can be technical features in the software itself, but can also be process elements in the development life-cycle, infrastructure features, etc. If using attack trees, the logical relationship represented by each branch will be either an AND or an OR. Therefore, by mitigating against just one precondition on an AND branch, the parent and all connected leaf nodes can be marked as mitigated. However, all child nodes on an OR node must be prevented before the parent can be marked as mitigated. Regardless of threat modeling technique, identify compensating controls and annotate the threat models directly. The goal is to maximize coverage in terms of controls that mark parts of the threat model as mitigated. For any viable paths remaining, identify potential compensating controls for feedback into organizational strategy. After initially conducted for a project, this must be updated and reviewed during the design phase or every development cycle. This activity should be conducted by a security auditor with relevant technical and business stakeholders.",
      "objective"  : "Concretely tie compensating controls to each threat against internal and third-party software",
      "proof"      : "Are the majority of the protection mechanisms and controls captured and mapped back to threats?"
    },
    "SR.1.A": {
      "description": "Conduct a review of functional requirements that specify the business logic and overall behavior for each software project. After gathering requirements for a project, conduct an assessment to derive relevant security requirements. Even if software is being built by a third-party, these requirements, once identified, should be included with functional requirements delivered to vendors. For each functional requirement, a security auditor should lead stakeholders through the process of explicitly noting any expectations with regard to security. Typically, questions to clarify for each requirement include expectations for data security, access control, transaction integrity, criticality of business function, separation of duties, uptime, etc. It is important to ensure that all security requirements follow the same principles for writing good requirements in general. Specifically, they should be specific, measurable, and reasonable. Conduct this process for all new requirements on active projects. For existing features, it is recommended to conduct the same process as a gap analysis to fuel future refactoring for security.",
      "objective"  : "Consider security explicitly during the software requirements process",
      "proof"      : "Do project teams specify security requirements during development?"
    },
    "SR.1.B": {
      "description": "Determine industry best-practices that project teams should treat as requirements. These can be chosen from publicly available guidelines, internal or external guidelines/standards/policies, or established compliance requirements. It is important to not attempt to bring in too many best-practice requirements into each development iteration since there is a time trade-off with design and implementation. The recommended approach is to slowly add best-practices over successive development cycles to bolster the software’s overall assurance profile over time. For existing systems, refactoring for security best practices can be a complex undertaking. Where possible, add security requirements opportunistically when adding new features. At a minimum, conducting the analysis to identify applicable best practices should be done to help fuel future planning efforts. This review should be performed by a security auditor with input from business stakeholders. Senior developers, architects, and other technical stakeholders should also be involved to bring design and implementation-specific knowledge into the decision process.",
      "objective"  : "Consider security explicitly during the software requirements process",
      "proof"      : "Do project teams pull requirements from best practices and compliance guidance?"
    },
    "SR.2.A": {
      "description": "ased upon the business purpose of the application, identify user and operator roles. Additionally, build a list of resources and capabilities by gathering all relevant data assets and application-specific features that are guarded by any form of access control. In a simple matrix with roles on one axis and resources on the other, consider the relationships between each role and each resource and note in each intersection the correct behavior of the system in terms of access control according to stakeholders. For data resources, it is important to note access rights in terms of creation, read access, update, and deletion. For resources that are features, gradation of access rights will likely be application-specific, but at a minimum note if the role should be permitted access to the feature. This permission matrix will serve as an artifact to document the correct access control rights for the business logic of the overall system. As such, it should be created by the project teams with input from business stakeholders. After initial creation, it should be updated by business stakeholders before every release, but usually toward the beginning of the design phase.",
      "objective"  : "Increase granularity of security requirements derived from business logic and known risks",
      "proof"      : "Do stakeholders review access control matrices for relevant projects?"
    },
    "SR.2.B": {
      "description": "Explicitly review existing artifacts that indicate organization or project-specific security risk in order to better understand the overall risk profile for the software. When available, draw on resources such as the high-level business risk profile, individual application threat models, findings from design review, code review, security testing, etc. In addition to review of existing artifacts, use abuse-case models for an application to serve as fuel for identification of concrete security requirements that directly or indirectly mitigate the abuse scenarios. This process should be conducted by business owners and security auditors as needed. Ultimately, the notion of risks leading to new security requirements should become a built-in step in the planning phase whereby newly discovered risks are specifically assessed by project teams.",
      "objective"  : "Increase granularity of security requirements derived from business logic and known risks",
      "proof"      : "Do project teams specify requirements based on feedback from other security activities?"
    },
    "SR.3.A": {
      "description": "Beyond the kinds of security requirements already identified by previous analysis, additional security benefits can be derived from third-party agreements. Typically, requirements and perhaps high-level design will be developed internally while detailed design and implementation is often left up to suppliers. Based on the specific division of labor for each externally developed component, identify specific security activities and technical assessment criteria to add to the vendor contracts. Commonly, this is a set of activities from the Design Review, Code Review, and Security Testing Practices. Modifications of agreement language should be handled on a case-by-case basis with each supplier since adding additional requirements will generally mean an increase in cost. The cost of each potential security activity should be balanced against the benefit of the activity as per the usage of the component or system being considered.",
      "objective"  : "Mandate security requirements process for all software projects and third-party dependencies",
      "proof"      : "Do stakeholders review vendor agreements for security requirements?"
    },
    "SR.3.B": {
      "description": "Incorporate checks for completeness of security requirements into routine project audits. Since this can be difficult to gauge without project-specific knowledge, the audit should focus on checking project artifacts such as requirements or design documentation for evidence that the proper types of analysis were conducted. Particularly, each functional requirement should be annotated with security requirements based on business drivers as well as expected abuse scenarios. The overall project requirements should contain a list of requirements generated from best-practices in guidelines and standards. Additionally, there should be a clear list of unfulfilled security requirements and an estimated timeline for their provision in future releases. This audit should be performed during every development iteration, ideally toward the end of the requirements process, but it must be performed before a release can be made.",
      "objective"  : "Mandate security requirements process for all software projects and third-party dependencies",
      "proof"      : "Are audits performed against the security requirements specified by project teams?"
    },
    "SA.1.A": {
      "description": "Across software projects within the organization identify commonly used third-party software libraries and frameworks in use. Generally, this need not be an exhaustive search for dependencies, but rather focus on capturing the high-level components that are most often used. From the list of components, group them into functional categories based on the core features provided by the third-party component. Also, note the usage prevalence of each component across project teams to weight the reliance upon the third-party code. Using this weighted list as a guide, create a list of components to be advertised across the development organization as recommended components. Several factors should contribute to decisions for inclusion on the recommended list. Although a list can be created without conducting research specifically, it is advisable to inspect each for incident history, track record for responding to vulnerabilities, appropriateness of functionality for the organization, excessive complexity in usage of the third-party component, etc. This list should be created by senior developers and architects, but also include input from managers and security auditors.  After creation, this list of recommended components matched against functional categories should be advertised to the development organization. Ultimately, the goal is to provide well-known defaults for project teams.",
      "objective"  : "Insert consideration of proactive security guidance into the software design process",
      "proof"      : "Are project teams provided with a list of recommended third-party components?"
    },
    "SA.1.B": {
      "description": "During design, technical staff on the project team should use a short list of guiding security principles as a checklist against detailed system designs. Typically, security principles include defense in depth, securing the weakest link, use of secure defaults, simplicity in design of security functionality, secure failure, balance of security and usability, running with least privilege, avoidance of security by obscurity, etc. In particular for perimeter interfaces, the design team should consider each principle in the context of the overall system and identify features that can be added to bolster security at each such interface. Generally, these should be limited such that they only take a small amount of extra effort beyond the normal implementation cost of functional requirements and anything larger should be noted and scheduled for future releases. While this process should be conducted by each project team after being trained with security awareness, it is helpful to incorporate more security-savvy staff to aide in making design decisions.",
      "objective"  : "Insert consideration of proactive security guidance into the software design process",
      "proof"      : "Are project teams aware of secure design principles and do they apply them consistently?"
    },
    "SA.2.A": {
      "description": "Organizations should identify shared infrastructure or services with security functionality. These will typically include single-sign-on services, corporate directory systems, access control or entitlements services, and authentication systems. By collecting and evaluating reusable systems, assemble a list of such resources and categorize them by the security mechanism they fulfill. It is also helpful to consider each resource in terms of why a development team would want to integrate with it, i.e. the benefits of using the shared resource. If multiple resources exist in each category, an organization should select and standardize on one or more shared service per category. Because future software development will rely on these selected services, each should be thoroughly audited to ensure the baseline security posture is understood. For each selected service, design guidance should be created for development teams to understand how to integrate with the system. After such guidance is assembled, it should be made available to development teams through training, mentorship, guidelines, and standards. The benefits of doing this include promotion of known-secure systems, simplified security guidance for project design teams, and clearer paths to building assurance around the applications utilizing the shared security services.",
      "objective"  : "Direct the software design process toward known-secure services and secure-by-default designs",
      "proof"      : "Do you advertise shared security services with guidance for project teams?"
    },
    "SA.2.B": {
      "description": "Across software projects at an organization, each should be categorized in terms of the generic architecture type. Common categories include client-server applications, embedded systems, desktop applications, web-facing applications, web services platforms, transactional middleware systems, mainframe applications, etc. Depending on your organizations specialty, more detailed categories may need to be developed based upon language, or processor architecture, or even era of deployment. For the generic software architecture type, a set of general design patterns representing sound methods of implementing security functionality can be derived and applied to the individual designs of an organization’s software projects. These security design patterns represent general definitions of generic design elements they can be researched or purchased, and it is often even more effective if these patterns are customized to be made more specific to your organization. Example patterns include a single-sign-on subsystem, a cross-tier delegation model, a hardened interface design, separation-of-duties authorization model, a centralized logging pattern, etc. The process of identification of applicable and appropriate patterns should be carried out by architects, senior developers, and other technical stakeholders during the design phase.",
      "objective"  : "Direct the software design process toward known-secure services and secure-by-default designs",
      "proof"      : "Are project teams provided with prescriptive design patterns based on their application architecture?"
    },
    "SA.3.A": {
      "description": "After promoting integration with shared security services and working with security patterns specific to each type of architecture, a collection of code implementing these pieces of functionality should be selected from project teams and used as the basis for a shared code-base. This shared code-base can initially start as a collection of commonly recommended libraries that each project needs to use and it can grow over time into one or more software frameworks representing reference platforms upon which project teams build their software. Examples of reference platforms include frameworks for model-view-controller web applications, libraries supporting transactional back-end systems, frameworks for web services platforms, scaffolding for client-server applications, frameworks for middle-ware with pluggable business logic, etc. Another method of building initial reference platforms is to select a particular project early in the life-cycle and have security-savvy staff work with them to build the security functionality in a generic way so that it could be extracted from the project and utilized elsewhere in the organization. Regardless of approach to creation, reference platforms have advantages in terms of speeding audit and security-related reviews, increasing efficiency in development, and lowering maintenance overhead. Architects, senior developers and other technical stakeholders should participate in design and creation of reference platforms. After creation, a team must maintain ongoing support and updates.",
      "objective"  : "Formally control the software design process and validate utilization of secure components",
      "proof"      : "Do project teams build software from centrally-controlled platforms and frameworks?"
    },
    "SA.3.B": {
      "description": "During routine audits of projects conduct additional analysis of project artifacts to measure usage of recommended frameworks, design patterns, shared security services, and reference platforms. Though conducted during routine audits, the goal of this activity is to collect feedback from project teams as much as to measure their individual proactive security effort. Overall, it is important to verify several factors with project teams. Identify use of non-recommended frameworks to determine if there may be a gap in recommendations versus the organization’s functionality needs. Examine unused or incorrectly used design patterns and reference platform modules to determine if updates are needed. Additionally, there may be more or different functionality that project teams would like to see implemented in the reference platforms as the organization evolves. This analysis can be conducted by any security-savvy technical staff. Metrics collected from each project should be collated for analysis by managers and stakeholders.",
      "objective"  : "Formally control the software design process and validate utilization of secure components",
      "proof"      : "Are project teams audited for the use of secure architecture components?"
    },
    "DR.1.A": {
      "description": "For each software project, create a simplified view of the overall architecture. Typically, this should be created based on project artifacts such as high-level requirements and design documents, interviews with technical staff, or module-level review of the code base. It is important to capture the high-level modules in the system, but a good rule of thumb for granularity is to ensure that the diagram of the whole system under review fits onto one page. From the single page architecture view, analyze each component in terms of accessibility of the interfaces from authorized users, anonymous users, operators, application-specific roles, etc. The components providing the interfaces should also be considered in the context of the one-page view to find points of functional delegation or data pass-through to other components on the diagram. Group interfaces and components with similar accessibility profiles and capture this as the software attack surface. For each interface, further elaborate the one-page diagram to note any security-related functionality. Based on the identified interface groups comprising the attack surface, check the model for design-level consistency for how interfaces with similar access are secured. Any breaks in consistency can be noted as assessment findings This analysis should be conducted by security-savvy technical staff, either within the project team or external. Typically, after initial creation, the diagram and attack surface analysis only needs to be updated during the design phase when additions or changes are made to the edge system interfaces.",
      "objective"  : "Support ad hoc reviews of software design to ensure baseline mitigations for known risks",
      "proof"      : "Do project teams document the attack perimeter of software designs?"
    },
    "DR.1.B": {
      "description": "Security requirements, either formally identified or informally known, should be identified and collected. Additionally, identify and include any security assumptions upon which safe operation of the system relies. Review each item on the list of known security requirements against the one-page diagram of the system architecture. Elaborate the diagram to show the design-level features that address each security requirement. Separate, granular diagrams can be created to simplify capturing this information if the system is large and/or complex. The overall goal is to verify that each known security requirement has been addressed by the system design. Any security requirements that are not clearly provided at the design level should be noted as assessment findings. This analysis should be conducted by security-savvy technical staff with input from architects, developers, managers, and business owners as needed.  It should be updated during the design phase when there are changes in security requirements or high-level system design.",
      "objective"  : "Support ad hoc reviews of software design to ensure baseline mitigations for known risks",
      "proof"      : "Do project teams check software designs against known security risks?"
    },
    "DR.2.A": {
      "description": "For each interface on a module in the high-level architecture diagram, formally iterate through the list of security mechanisms and analyze the system for their provision. This type of analysis should be performed on both internal interfaces, e.g. between tiers, as well as external ones, e.g. those comprising the attack surface. The six main security mechanisms to consider are authentication, authorization, input validation, output encoding, error handling and logging. Where relevant, also consider the mechanisms of cryptography and session management. For each interface, determine where in the system design each mechanism is provided and note any missing or unclear features as findings. This analysis should be conducted by security-savvy staff with assistance from the project team for application-specific knowledge. This analysis should be performed once per release, usually toward the end of the design phase. After initial analysis, subsequent releases are required to update the findings based on changes being made during the development cycle.",
      "objective"  : "Offer assessment services to review software design against comprehensive best practices for security",
      "proof"      : "Do project teams specifically analyze design elements for security mechanisms?"
    },
    "DR.2.B": {
      "description": "Institute a process whereby project stakeholders can request an design review. This service may be provided centrally within the organization or distributed across existing staff, but all reviewers must be trained on performing the reviews completely and consistently. The review service should be centrally managed in that the review request queue should be triaged by senior managers, architects, and stakeholders that are familiar with the overall business risk profile for the organization. This allows prioritization of project reviews in alignment with overall business risk. During a design review, the review team should work with project teams to collect information sufficient to formulate an understanding of the attack surface, match project-specific security requirements to design elements, and verify security mechanisms at module interfaces.",
      "objective"  : "Offer assessment services to review software design against comprehensive best practices for security",
      "proof"      : "Are project stakeholders aware of how to obtain a formal secure design review?"
    },
    "DR.3.A": {
      "description": "Based on the business function of the software project, conduct analysis to identify details on system behavior around high-risk functionality. Typically, high-risk functionality will correlate to features implementing creation, access, update, and deletion of sensitive data. Beyond data, high-risk functionality also includes project-specific business logic that is critical in nature, either from a denial-of-service or compromise perspective. For each identified data source or business function, select and use a standardized notation to capture relevant software modules, data sources, actors, and messages that flow amongst them. It is often helpful to start with a high-level design diagram and iteratively flesh out relevant detail while removing elements that do not correspond to the sensitive resource. With data-flow diagrams created for a project, conduct analysis over them to determine internal choke-points in the design. Generally, these will be individual software modules that handle data with differing sensitivity levels or those that gate access to several business functions of various levels of business criticality.",
      "objective"  : "Require assessments and validate artifacts to develop detailed understanding of protection mechanisms",
      "proof"      : "Does the secure design review process incorporate detailed data-level analysis?"
    },
    "DR.3.B": {
      "description": "Having established a consistent design review program, the next step of enforcement is to set a particular point in the software development life-cycle where a project cannot pass until an design review is conducted and findings are reviewed and accepted. In order to accomplish this, a baseline level of expectations should be set, e.g. no projects with any high-severity findings will be allowed to pass and all other findings must be accepted by the business owner. Generally, design reviews should occur toward the end of the design phase to aide early detection of security issues, but it must occur before releases can be made from the project team. For legacy systems or inactive projects, an exception process should be created to allow those projects to continue operations, but with an explicitly assigned time-frame for each to be reviewed to illuminate any hidden vulnerabilities in the existing systems. Exceptions for should be limited to no more than 20% of all projects.",
      "objective"  : "Require assessments and validate artifacts to develop detailed understanding of protection mechanisms",
      "proof"      : "Does a minimum security baseline exist for secure design review results?"
    },
    "IR.1.A": {
      "description": "From the known security requirements for a project, derive a lightweight code review checklist for security. These can be checks specific to the security concerns surrounding the functional requirements or checks for secure coding best practices based on the implementation language, platform, typical technology stack, etc. Due to these variations, often a set of checklist are needed to cover the different types of software development within an organization. Regardless, of whether created from publicly available resources or purchased, technical stakeholders such as development managers, architects, developers, and security auditors should review the checklists for efficacy and feasibility. It is important to keep the lists short and simple, aiming to catch high-priority issues that are straightforward to find in code either manually or with simple search tools. Code analysis automation tools may also be used to achieve this same end, but should also be customized to reduce the overall set of security checks to a small, valuable set in order to make the scan and review process efficient. Developers should be briefed on the goals of checklists appropriate to their job function.",
      "objective"  : "Opportunistically find basic code-level vulnerabilities and other high-risk security issues",
      "proof"      : "Do project teams have review checklists based on common security related problems?"
    },
    "IR.1.B": {
      "description": "Since code-level vulnerabilities can have dramatically increased impacts if they occur in security-critical parts of software, project teams should review high-risk modules for common vulnerabilities. Common examples of high-risk functionality include authentication modules, access control enforcement points, session management schemes, external interfaces, input validators and data parsers, etc. Utilizing the code review checklists, the analysis can be performed as a normal part of the development process where members of the project team are assigned modules to review when changes are made. Security auditors and automated review tools can also be utilized for the review. During development cycles where high-risk code is being changed and reviewed, development managers should triage the findings and prioritize remediation appropriately with input from other project stakeholders.",
      "objective"  : "Opportunistically find basic code-level vulnerabilities and other high-risk security issues",
      "proof"      : "Do project teams review selected high-risk code?"
    },
    "IR.2.A": {
      "description": "Many security vulnerabilities at the code level are complex to understand  and require careful inspection for discovery. However, there are many useful automation solutions available to automatically analyze code for bugs and vulnerabilities. There are both commercial and open-source products available to cover popular programming languages and frameworks. Selection of an appropriate code analysis solution is based on several factors including depth and accuracy of inspection, product usability and usage model, expandability and customization features, applicability to the organization’s architecture and technology stack(s), etc. Utilize input from security-savvy technical staff as well as developers and development managers in the selection process, and review overall results with stakeholders.",
      "objective"  : "Make code review during development more accurate and efficient through automation",
      "proof"      : "Can project teams access automated code analysis tools to find security problems?"
    },
    "IR.2.B": {
      "description": "Once a code analysis solution is selected, it must be integrated into the development process to encourage project teams to utilize its capabilities. An effective way to accomplish this is to setup the infrastructure for the scans to run automatically at build time or from code in the project’s code repository. In this fashion, results are available earlier thus enabling development teams to self-check along the way before release. A potential problem with legacy systems or large ongoing projects is that code scanners will typically report findings in modules that were not being updated in the release. If automatic scanning is setup to run periodically, an effective strategy to avoid review overhead is to limit consideration of findings to those that have been added, removed, or changed since the previous scan. If is critical to not ignore the rest of the results however, so development managers should take input from security auditors, stakeholders, and the project team to formulate a concrete plan for addressing the rest of the findings. If unaddressed findings from code review remain at release, these must be reviewed and accepted by project stakeholders.",
      "objective"  : "Make code review during development more accurate and efficient through automation",
      "proof"      : "Do stakeholders consistently review results from code reviews?"
    },
    "IR.3.A": {
      "description": "Code scanning tools are powered by built-in a knowledge-base of rules to check code based on language APIs and commonly used libraries, but have limited ability to understand custom APIs and designs to apply analogous checks. However, through customization, a code scanner can be a powerful, generic analysis engine for finding organization and project-specific security concerns.  While details vary between tools in terms of ease and power of custom analysis, code scanner customization generally involves specifying checks to be performed at specific APIs and function call sites. Checks can include analysis for adherence to internal coding standards, unchecked tainted data being passed to custom interfaces, tracking and verification of sensitive data handling, correct usage of an internal API, etc. Checkers for usage of shared code-bases are an effective place to begin scanner customizations since the created checkers can be utilized across multiple projects. To customize a tool for a code-base, a security auditor should inspect both code and high-level design to identify candidate checkers to discuss with development staff and stakeholders for implementation.",
      "objective"  : "Mandate comprehensive code review process to discover language-level and application-specific risks",
      "proof"      : "Do project teams utilize automation to check code against application-specific coding standards?"
    },
    "IR.3.B": {
      "description": "To set a code-level security baseline for all software projects, a particular point in the software development life-cycle should be established as a checkpoint where a minimum standard for code review results must be met in order to make a release.  To begin, this standard should be straightforward to meet, for example by choosing one or two vulnerability types and a setting the standard that no project may pass with any corresponding findings. Over time, this baseline standard should be improved by adding additional criteria for passing the checkpoint. Generally, the code review checkpoint should occur toward the end of the implementation phase, but must occur before release.  For legacy systems or inactive projects, an exception process should be created to allow those projects to continue operations, but with an explicitly assigned timeframe for mitigation of findings. Exceptions should be limited to no more that 20% of all projects.",
      "objective"  : "Mandate comprehensive code review process to discover language-level and application-specific risks",
      "proof"      : "Does a minimum security baseline exist for code review results?"
    },
    "ST.1.A": {
      "description": "From the known security requirements for a project, identify a set of test cases to check the software for correct functionality.  Typically, these test cases are derived from security concerns surrounding the functional requirements and business logic of the system, but should also include generic tests for common vulnerabilities based on the implementation language or technology stack. Often, it is most effective to use the project team’s time to build application-specific test cases and utilize publicly available resources or purchased knowledge bases to select applicable general test cases for security. Although not required, automated security testing tools can also be utilized to cover the general security test cases. This test case planning should occur during the requirements and/or design phases, but must occur before final testing prior to release. Candidate test cases should be reviewed for applicability, efficacy, and feasibility by relevant development, security, and quality assurance staff.",
      "objective"  : "Establish process to perform basic security tests based on implementation and software requirements",
      "proof"      : "Do projects specify security testing based on defined security requirements?"
    },
    "ST.1.B": {
      "description": "Using the set of security test cases identified for each project, penetration testing should be conducted to evaluate the system’s performance against each case. It is common for this to occur during the testing phase prior to release. Penetration testing cases should include both application-specific tests to check soundness of business logic as well as common vulnerability tests to check the design and implementation. Once specified, security test cases can be executed by security-savvy quality assurance or development staff, but first-time execution of security test cases for a project team should be monitored by a security auditor to assist and coach team members. Prior to release or deployment, stakeholders must review results of security tests and accept the risks indicated by failing security tests at release time. In the latter case, a concrete timeline should be established to address the gaps over time.",
      "objective"  : "Establish process to perform basic security tests based on implementation and software requirements",
      "proof"      : "Is penetration testing performed on high risk projects prior to release?"
    },
    "ST.2.A": {
      "description": "In order to test for security issues, a potentially large number of input cases must be checked against each software interface, which can make effective security testing using manual test case implementation and execution unwieldy.  Thus, automated security test tools should be used to automatically test software, resulting in more efficient security testing and higher quality results. Both commercial and open-source products are available and should be reviewed for appropriateness for the organization. Selecting a a suitable tool is based on several factors including robustness and accuracy of built-in security test cases, efficacy at testing architecture types important to organization, customization to change or add test cases, quality and usability of findings to the development organization, etc.. Utilize input from security-savvy technical staff as well as development and quality assurance staff in the selection process, and review overall results with stakeholders.",
      "objective"  : "Make security testing during development more complete and efficient through automation",
      "proof"      : "Do projects use automation to evaluate security test cases?"
    },
    "ST.2.B": {
      "description": "With tools to run automated security tests, projects within the organization should routinely run security tests and review results during development. In order to make this scalable with low overhead, security testing tools should be configured to automatically run on a routine basis, e.g. nightly or weekly, and findings should be inspected as they occur.  Conducting security tests as early as the requirements or design phases can be beneficial. While traditionally, used for functional test cases,  this type of test-driven development approach involves identifying and running relevant security test cases early in the development cycle, usually during design. With the automatic execution of security test cases, projects enter the implementation phase with a number of failing tests for the non-existent functionality. Implementation is complete when all the tests pass.   This provides a clear, upfront goal for developers early in the development cycle, thus lowering risk of release delays due to security concerns or forced acceptance of risk in order to meet project deadlines. For each project release, results from automated and manual security tests should be presented to management and business stakeholders for review. If there are unaddressed findings that remain as accepted risks for the release, stakeholders and development managers should work together to establish a concrete timeframe for addressing them.",
      "objective"  : "Make security testing during development more complete and efficient through automation",
      "proof"      : "Do projects follow a consistent process to evaluate and report on security tests to stakeholders?"
    },
    "ST.3.A": {
      "description": "Through either customization of security testing tools, enhancements to generic test case execution tools, or buildout of custom test harnesses, project teams should formally iterate through security requirements and build a set of automated checkers to test the security of the implemented business logic. Additionally, many automated security testing tools can be greatly improved in accuracy and depth of coverage if they are customized to understand more detail about the specific software interfaces in the project under test. Further, organization-specific concerns from compliance or technical standards can be codified as a reusable, central test battery to make audit data collection and per-project management visibility simpler. Project teams should focus on buildout of granular security test cases based on the business functionality of their software, and an organization-level team led by a security auditor should focus on specification of automated tests for compliance and internal standards.",
      "objective"  : "Require application-specific security testing to ensure baseline security before deployment",
      "proof"      : "Are security test cases comprehensively generated for application-specific logic?"
    },
    "ST.3.B": {
      "description": "To prevent software from being released with easily found security bugs, a particular point in the software development life-cycle should be identified as a checkpoint where an established set of security test cases must pass in order to make a release from the project.  This establishes a baseline for the kinds of security tests all projects are expected to pass. Since adding too many test cases initially can result in an overhead cost bubble, begin by choosing one or two security issues and include a wide variety of test cases for each with the expectation that no project may pass if any test fails. Over time, this baseline should be improved by selecting additional security issues and adding a variety of corresponding test cases. Generally, this security testing checkpoint should occur toward the end of the implementation or testing, but must occur before release.  For legacy systems or inactive projects, an exception process should be created to allow those projects to continue operations, but with an explicitly assigned timeframe for mitigation of findings. Exceptions should be limited to no more that 20% of all projects.",
      "objective"  : "Require application-specific security testing to ensure baseline security before deployment",
      "proof"      : "Does a minimum security baseline exist for security testing? "
    },
    "IM.1.A": {
      "description": "For each division within the organization or for each project team, establish a point of contact to serve as a communications hub for security information. While generally this responsibility will not claim much time from the individuals, the purpose of having a predetermined point of contact is to add structure and governance for vulnerability management. Examples of incidents that might cause the utilization include receipt of a vulnerability report from an external entity, compromise or other security failure of software in the field,  internal discovery of high-risk vulnerabilities, etc. In case of an event, the closest contact would step in as an extra resource and advisor to the affected project team(s) to provide technical guidance and brief other stakeholders on progress of mitigation efforts.   The point of contact should be chosen from security-savvy technical or management staff with a breadth of knowledge over the software projects in the organization. A list of these assigned security points of contact should be centrally maintained and updated at least every six months. Additionally, publishing and advertising this list allows staff within the organization to request help and work directly with one another on security problems.",
      "objective"  : "Understand high-level plan for responding to vulnerability reports or incidents",
      "proof"      : "Do projects have a point of contact for security issues or incidents?"
    },
    "IM.1.B": {
      "description": "From the list of individuals assigned responsibility as a security point of contact or from dedicated security personnel, select a small group to serve as a centralized technical security response team. The responsibilities of the team will include directly taking ownership of security incidents or  vulnerability reports and being responsible for triage, mitigation, and reporting to stakeholders. Given their responsibility when tapped, members of the security response team are also responsible for executive briefings and upward communication during an incident. It is likely that most of the time, the security response team would not be operating in this capacity, though they must be flexible enough to be able to respond quickly or a smooth process must exist for deferring and incident to another team member. The response team should hold a meeting at least annually to brief security points of contact on the response process and high-level expectations for security-related reporting from project teams.",
      "objective"  : "Understand high-level plan for responding to vulnerability reports or incidents",
      "proof"      : "Does your organization have an assigned security response team?"
    },
    "IM.2.A": {
      "description": "Extending from the informal security response team, explicitly document  the organization’s incident response process as well as the procedures that  team members are expected to follow.  Additionally, each member of the security response team must be trained on this material at least annually. There are several tenets to sound incident response process and they include initial triage to prevent additional damage, change management and patch application, managing project personnel and others involved in the incident, forensic evidence collection and preservation, limiting communication about the incident to stakeholders, well-defined reporting to stakeholders and/or communications trees, etc. With development teams, the security responders should work together to conduct the technical analysis to verify facts and assumptions about each incident or vulnerability report. Likewise, when project teams detect an incident or high-risk vulnerability, they should follow an internal process that puts them in contact with a member of the security response team.",
      "objective"  : "Elaborate expectations for response process to improve consistency and communications",
      "proof"      : "Does the organization utilize a consistent process for incident reporting and handling?"
    },
    "IM.2.B": {
      "description": "For most organizations, it is undesirable to let news of a security problem become public, but there are several important ways in which internal-to-external communications on security issues should be fulfilled. The first and most common is through creation and deployment of security patches for the software produced by the organization. Generally, if all software projects are only used internally, then this becomes less critical, but for all contexts where the software is being operated by parties external to the organization, a patch release process must exist. It should provide for several factors including change management and regression testing prior to patch release, announcement to operators/users with assigned criticality category for the patch, sparse technical details so that an exploit cannot be directly derived, etc. Another avenue for external communications is with third parties that report security vulnerabilities in an organization’s software. By adopting and externally posting the expected process with timeframes for response, vulnerability reporters are encouraged to follow responsible disclosure practices. Lastly, many states and countries legally require external communications for incidents involving data theft of personally identifiable information and other sensitive data type. Should this type of incident occur, the security response team should work with managers and business stakeholders to determine appropriate next-steps.",
      "objective"  : "Elaborate expectations for response process to improve consistency and communications",
      "proof"      : "Are project stakeholders aware of relevant security disclosures related to their software projects?"
    },
    "IM.3.A": {
      "description": "Though potentially time consuming, the incident response process should be augmented to include additional analysis to identify the key, underlying security failures. These root causes can be technical problems such as code-level vulnerabilities, configuration errors, etc. or they can be people/process problems such as social engineering, failure to follow procedures, etc. Once a root cause is identified for an incident, it should be used as a tool to find other potential weaknesses in the organization where an analogous incident could have occurred. For each identified weakness additional recommendations for proactive mitigations should be communicated as part of closing out the original incident response effort. Any recommendations based on root cause analysis should be reviewed by management and relevant business stakeholders in order to either schedule mitigation activities or note the accepted risks.",
      "objective"  : "Improve analysis and data gathering within response process for feedback into proactive planning",
      "proof"      : "Are incidents inspected for root causes to generate further recommendations?"
    },
    "IM.3.B": {
      "description": "By having a centralized process to handle all compromise and high-priority vulnerability reports, an organization is enabled to take measurements of trends over time to determine impact and efficiency of initiatives for security assurance. Records of past incidents should be stored and reviewed at least every 6 months. Group similar incidents and simply tally the overall count for each type of problem.  Additional measurements to take from the incidents include frequency of software projects affected by incidents, system downtime and cost from loss of use, human resources taken in handling and cleanup of the incident, estimates of long-term costs such as regulatory fines or brand damage, etc.  For root causes that were technical problems in nature, it is also helpful to identify what kind of proactive, review, or operational practice might have detected it earlier or lessened the damage. This information is concrete feedback into the program planning process since it represents the real security impact that the organization has felt over time.",
      "objective"  : "Improve analysis and data gathering within response process for feedback into proactive planning",
      "proof"      : "Do projects consistently collect and report data and metrics related to incidents?"
    },
    "EH.1.A": {
      "description": "For each project, a concrete definition of the expected operating platforms should be created and maintained. Depending on the organization, this specification should be jointly created with development staff, stakeholders, support and operations groups, etc. Begin this specification should by capturing all details that must be true about the operating environment based upon the business function of the software. These can include factors such as processor architecture, operating system versions, prerequisite software, conflicting software, etc. Further, note any known user or operator configurable options about the operating environment that affect the way in which the software will behave. Additionally, identify any relevant assumptions about the operating environment that were made in design and implementation of the project and capture those assumptions in the specification. This specification should be reviewed and updated at least every 6 months for active projects or more often if changes are being made to the software design or the expected operating environment.",
      "objective"  : "Understand baseline operational environment for applications and software components",
      "proof"      : "Do projects document operational environment security requirements?"
    },
    "EH.1.B": {
      "description": "Most applications are software that runs on top of another large stack of software composed of built-in programming language libraries, third-party components and development frameworks, base operating systems, etc. Because security flaws contained in any module in that large software stack affect the overall security of the organization’s software, critical security updates for elements of the technology stack must be installed. As such, regular research or ongoing monitoring of high-risk dependencies should be performed to stay abreast of the latest fixes to security flaws. Upon identification of a critical upgrade or patch that would impact the security posture of the software project, plans should be made to get affected users and operators to update their installations. Depending on the type of software project, details on doing this can vary.",
      "objective"  : "Understand baseline operational environment for applications and software components",
      "proof"      : "Do projects check for security updates to third-party software components?"
    },
    "EH.2.A": {
      "description": "Moving to a more formal process than ad hoc application of critical upgrades and patches, an ongoing process should be created in the organization to consistently apply updates to software dependencies in the operating environment. In the most basic form, the process should aim to make guarantees for time lapse between release and application of security upgrades and patches. To make this process efficient, organizations typically accept high latency on lower priority updates, e.g. maximum of 2 days for critical patches spanning to a maximum of 30 days for low priority patches. This activity should be primarily conducted by support and operations staff, but routine meetings with development should also be conducted to keep the whole project abreast of past changes and scheduled upgrades.  Additionally, development staff should share a list of third-party components upon which the software project internally depends so that support and operations staff can monitor those as well to cue development teams on when an upgrade is required.",
      "objective"  : "Improve confidence in application operations by hardening the operating environment",
      "proof"      : "Is a consistent process used to apply upgrades and patches to critical dependencies?"
    },
    "EH.2.B": {
      "description": "Given the complexity of monitoring and managing patches alone across the variety of components composing the infrastructure for a software project, automation tools should be utilized to automatically monitor systems for soundness of configuration. There are both commercial and open-source tools available to provide this type of functionality, so project teams should select a solution based on appropriateness to the organization’s needs. Typical selection criteria includes ease of deployment and customization, applicability to the organization’s platforms and technology stacks, built-in features for change management and alerting, metrics collection and trend tracking etc. In addition to host and platform checks, monitoring automation should be customized to perform application-specific health checks and configuration verifications. Support and operations personnel should work with architects and developers to determine the optimal amount of monitoring for a given software project. Ultimately, after a solution is deployed for monitoring the environment’s configuration status, unexpected alerts or configuration changes should be collected and regularly reviewed by project stakeholders as often as weekly but at least once per quarter.",
      "objective"  : "Improve confidence in application operations by hardening the operating environment",
      "proof"      : "Do projects leverage automation to check application and environment health?"
    },
    "EH.3.A": {
      "description": "In order to build a better assurance case for software in its operating environment, additional tools can be used to enhance the security posture of the overall system. Operational environments can vary dramatically, thus the appropriateness of given protection technology should be considered in the project context. Commonly used protections tools include web application firewalls, XML security gateways for web services, anti-tamper and obfuscation packages for client/embedded systems, network intrusion detection/prevention systems for legacy infrastructure, forensic log aggregation tools, host-based integrity verification tools, etc. Based on the organization and project-specific knowledge, technical stakeholders should work with support and operations staff to identify and recommend selected operations protection tools to business stakeholders. If deemed a valuable investment in terms of risk-reduction versus cost of implementation, stakeholders should agree on plans for a pilot, widespread rollout, and ongoing maintenance.",
      "objective"  : "Validate application health and status of operational environment against known best practices",
      "proof"      : "Are stakeholders aware of options for additional tools to protect software while running in operations?"
    },
    "EH.3.B": {
      "description": "When conducting routine project-level audits, expand the review to include inspection of artifacts related to hardening the operating environment. Beyond an up-to-date specification for the operational environment, audits should inspect current patch status and historic data since the previous audit. By tapping into monitoring tools, audits can also verify key factors about application configuration management and historic changes. Audits should also inspect the usage of operations protections tools against those available for the software’s architecture type. Audits for infrastructure can occur at any point after a project’s initial release and deployment, but should occur at least every 6 months. For legacy systems or projects without active development, infrastructure audits should still be conducted and reviewed by business stakeholders. An exception process should be created to allow special-case projects to continue operations, but with an explicitly assigned timeframe for mitigation of findings. Exceptions should be limited to no more that 20% of all projects.",
      "objective"  : "Validate application health and status of operational environment against known best practices",
      "proof"      : "Does a minimum security baseline exist for environment health (versioning, patching, etc)?"
    },
    "OE.1.A": {
      "description": "With software-specific knowledge, project teams should identify any security-relevant configuration and operations information and communicate it to users and operators. This enables the actual security posture of software at deployment sites to function in the same way that designers in the project team intended.  This analysis should begin with architects and developers building a list of security features built-in to the software. From that list, information about configuration options and their security impact should be captured as well. For projects that offer several different deployment models, information about the security ramifications of each should be noted to better inform users and operators about the impact of their choices. Overall, the list should be lightweight and aim to capture the most critical information. Once initially created, it should be reviewed by the project team and business stakeholders for agreement. Additionally, it is effective to review this list with select operators or users in order to ensure the information is understandable and actionable. Project teams should review and update this information with every release, but must do so at least every 6 months.",
      "objective"  : "Enable communications between development teams and operators for critical security-relevant data",
      "proof"      : "Are security notes delivered with each software release?"
    },
    "OE.1.B": {
      "description": "With specific knowledge of ways in which software behaves, project teams should identify the most important error and alert messages which require user/operator attention. From each identified event, information related to appropriate user/operator actions in response to the event should be captured. From the potentially large set of events that the software might generate, select the highest priority set based on relevance in terms of the business purpose of the software. This should include any security-related events, but also may include critical errors and alerts related to software health and configuration status. For each event, actionable advice should be captured to inform users and operators of required next steps and potential root causes of the event. These procedures must be reviewed by the project team and updated at every major product release, every 6 months, but can be done more frequently, e.g. with each release.",
      "objective"  : "Enable communications between development teams and operators for critical security-relevant data",
      "proof"      : "Are security-related alerts and error conditions documented on a per-project basis?"
    },
    "OE.2.A": {
      "description": "To more formally update users and operators on relevant changes in the software, each release must include change management procedures relevant to upgrade and first-time installation. Overall, the goal is to capture the expected accompanying steps that ensure the deployment will be successful and not incur excessive downtime or degradation of security posture. To build these procedures during development, the project teams should setup a lightweight internal process for capturing relevant items that would impact deployments. It is effective to have this process in place early in the development cycle so that this information can be retained as soon as it is identified while in the requirements, design, and implementation phases. Before each release, the project team should review the list as a whole for completeness and feasibility. For some projects, extensive change procedures accompanying a given release may warrant special handling, such as building automated upgrade scripts to prevent errors during deployment.",
      "objective"  : "Improve expectations for continuous secure operations through provision of detailed procedures",
      "proof"      : "Do projects utilize a change management process that’s well understood?"
    },
    "OE.2.B": {
      "description": "Starting from the information captured on critical software events and the procedures for handling each, project teams should build and maintain formal guides that capture all the security-relevant information that users and operators need to know. Initially, this guide should be built from the known information about the system, such as security-related configuration options, event handling procedures, installation and upgrade guides, operational environment specifications, security-related assumptions about the deployment environment, etc. Extending this, the formal operational security guide should elaborate on each of these to cover more details such that the majority of the users and operators will be informed for all the questions they might have had. For large or complex systems, this can be challenging, so project teams should work with business stakeholders to determine the appropriate level of documentation. Additionally, project teams should document any recommendations for deployments that would enhance security. The operational security guide, after initial creation, should be reviewed by project teams and updated with each release.",
      "objective"  : "Improve expectations for continuous secure operations through provision of detailed procedures",
      "proof"      : "Do project teams deliver an operational security guide with each product release?"
    },
    "OE.3.A": {
      "description": "When conducting routine project-level audits, expand the review to include inspection of artifacts related to operational enablement for security. Projects should be checked to ensure they have an updated and complete operational security guides as relevant to the specifics of the software. These audits should begin toward the end of the development cycle close to release, but must be completed and passed before a release can be made. For legacy systems or inactive projects, this type of audit should be conducted and a one-time effort should be made to address findings and verify audit compliance, after which additional audits for operational enablement are no longer required.  Audit results must be reviewed with business stakeholders prior to release. An exception process should be created to allow projects failing an audit to continue with a release, but these projects should have a concrete timeline for mitigation of findings. Exceptions should be limited to no more that 20% of all active projects.",
      "objective"  : "Mandate communication of security information and validate artifacts for completeness",
      "proof"      : "Are project releases audited for appropriate operational security information?"
    },
    "OE.3.B": {
      "description": "Though often used with special-purpose software, code signing allows users and operators to perform integrity checks on software such that they can cryptographically verify the authenticity of a module or release. By signing software modules, the project team enables deployments to operate with a greater degree of assurance against any corruption or modification of the deployed software in its operating environment. Signing code incurs overhead for management of signing credentials for the organization. An organization must follow safe key management processes to ensure the ongoing confidentiality of the signing keys. When dealing with any cryptographic keys, project stakeholders must also consider plans for dealing with common operational problems related to cryptography such as key rotation, key compromise, or key loss. Since code signing is not appropriate for everything, architects and developers should work with security auditors and business stakeholders to determine which parts of the software should be signed. As projects evolve, this list should be reviewed with each release, especially when adding new modules or making changes to previously signed components.",
      "objective"  : "Mandate communication of security information and validate artifacts for completeness",
      "proof"      : "Is code signing routinely performed on software components using a consistent process?"
    }

  }
}