draft — open for review

CVE DQAF — Canonical Task List

Purpose

This document collapses the 273 role-level tasks from the tasks-by-role working document into a canonical set of 45 distinct tasks. Each canonical task represents a discrete practitioner activity with coherent data quality requirements. Tasks that differed only in who performs them — not in what data they require — are merged. Tasks that appear similar but require materially different data are kept separate.

Each task lists primary roles (roles for whom this is a core, regular activity) and secondary roles (roles that perform this task occasionally or as a downstream consumer of its outputs).

Role numbers reference the role inventory in the Role Inventory document.


Role Reference (abbreviated)

# Role
1 Vulnerability finder / security researcher (internal)
2 PSIRT coordinator
3 CNA record author
4 Security engineer (producer-side)
5 Developer / fix author
6 Technical writer / advisory author
7 Customer support representative
8 Sales / account manager
9 Vulnerability analyst
10 Patch manager
11 SOC analyst
12 Threat intelligence analyst
13 Detection engineer
14 Application security engineer
15 Penetration tester / red teamer
16 Risk officer / CISO
17 Compliance officer
18 IT operations / system administrator
19 Vulnerability data engineer
20 Asset inventory / CMDB owner
21 Procurement / vendor risk analyst
22 Independent security researcher / finder
23 Exploit developer
24 Bug bounty participant
25 Security tool vendor / developer
26 SBOM tooling developer
27 Vulnerability enricher / ADP operator
28 Vulnerability database curator
29 Vulnerability database operator
30 Data pipeline engineer
31 Disambiguation / identity resolution analyst
32 Vulnerability feed aggregator / broker
33 Threat intelligence platform integrator
34 CVE program administrator / Root CNA
35 CNA quality reviewer
36 Standards body representative
37 Government / national CERT analyst
38 Policy maker / regulator
39 Academic / industry researcher
40 Data analyst (government or industry)
41 Journalist

Canonical Tasks

Phase 1: Pipeline and Data Management


T01. Ingest and normalize CVE records from primary sources

Fetch CVE records from cve.org, NVD, GHSA, OSV, vendor feeds, and other primary sources. Normalize format across schema versions (CVE JSON 4.0, 5.0), source variants, and field naming conventions into a consistent internal representation.

Role type Roles
Primary 19, 25, 27, 29, 30, 32
Secondary 9, 28

T02. Detect and propagate record changes and updates

Monitor upstream sources for changes to existing records. Identify which fields changed, when, and propagate updates to downstream consumers. Distinct from initial ingest — requires delta detection and change tracking.

Role type Roles
Primary 19, 25, 30, 32
Secondary 27, 29

T03. Monitor feed health and handle source disruptions

Track availability and freshness of upstream CVE data sources. Detect feed failures, delays, or unexpected schema changes and respond to maintain data currency downstream.

Role type Roles
Primary 19, 29, 30, 32
Secondary 25

T04. Deduplicate CVE records across sources

Identify and collapse duplicate or near-duplicate records that arrive from multiple sources referencing the same vulnerability. Distinct from identity resolution (T05), which handles cases requiring analytical judgment.

Role type Roles
Primary 19, 30, 31, 32
Secondary 28

T05. Resolve cross-source identity

Determine analytically whether records from different sources (CVE, GHSA, OSV, vendor advisory) describe the same underlying vulnerability. Requires domain knowledge and judgment. Produces crosswalk tables and alias mappings.

Role type Roles
Primary 28, 31, 32
Secondary 27, 30

T06. Resolve split and merge cases

Determine whether a single CVE ID covers multiple distinct vulnerabilities that should be separated (split), or whether multiple CVE IDs cover the same vulnerability and should be unified (merge). Requires program-level judgment and coordination.

Role type Roles
Primary 28, 31, 34
Secondary 27, 32

Phase 2: Asset-to-Vulnerability Matching


T07. Match CVE records to assets in an inventory or manifest

Determine which CVE records apply to software, hardware, or services present in a technology estate. May be automated (scanner, SBOM tool) or manual. The primary identification task — all downstream tasks depend on it.

Role type Roles
Primary 9, 14, 25, 26
Secondary 11, 15, 22

T08. Resolve false positives from identifier failures

Investigate and dismiss scanner findings where the CVE-to-asset match is incorrect — caused by CPE naming inconsistencies, wrong version range encoding, missing platform constraints, or cross-ecosystem identifier collision.

Role type Roles
Primary 9, 25
Secondary 14, 19

T09. Confirm match applicability for specific configurations or deployments

Determine whether a matched CVE actually applies given the organization’s specific configuration, deployment context, feature set, or platform. Requires reading the exploitation conditions and comparing them to local state.

Role type Roles
Primary 4, 9, 14
Secondary 1, 15

Phase 3: Record Consumption and Interpretation


T10. Look up and interpret a CVE record

Read a CVE record to understand what vulnerability it describes: what the fault is, what conditions allow exploitation, and what the security consequence is. The baseline task that most other tasks build on. Performed across virtually every role.

Role type Roles
Primary 9, 11, 12, 13, 15, 22, 23, 37
Secondary 7, 8, 16, 17, 21, 24, 38, 41

Phase 4: Vulnerability Discovery and Characterization


T11. Discover a vulnerability

Identify a previously unknown vulnerability through code review, fuzzing, dynamic analysis, or reverse engineering. The starting point of the producer-side lifecycle.

Role type Roles
Primary 1, 22, 24
Secondary 15

T12. Reproduce and root-cause a vulnerability

Confirm that a reported or discovered vulnerability is real and identify the specific code-level or design-level fault that causes it. Required for accurate CVE record authorship and fix development.

Role type Roles
Primary 1, 4, 22, 23
Secondary 13, 15, 27

T13. Enumerate exploitation conditions

Determine the complete set of circumstances that must hold for the vulnerability to be exploitable: authentication requirements, network access, required privileges, configuration dependencies, and preconditions. Higher specificity than CVSS attack metrics.

Role type Roles
Primary 1, 4, 13, 22
Secondary 15, 23

T14. Determine affected versions, configurations, and platforms

Identify exactly which versions of the affected product are vulnerable, which configurations trigger the vulnerability, and which platforms are in scope. Determines what the CVE affected range should contain.

Role type Roles
Primary 1, 3, 4, 22
Secondary 9, 14

Phase 5: CVE Record Production and Maintenance


T15. Author and publish a CVE record

Write and publish a CVE record covering description, CWE assignment, CVSS vector, affected product and version ranges, reference tags, and any optional fields (workarounds, solutions, timeline). The primary data production task.

Role type Roles
Primary 3
Secondary 2, 4, 27, 34

T16. Update a CVE record post-publication

Revise a published CVE record when new information becomes available: additional affected versions discovered, fix availability changes, CVSS corrected, dispute resolved, or enrichment superseded. Distinct from initial authorship.

Role type Roles
Primary 3, 27
Secondary 2, 34, 35

Phase 6: Disclosure Coordination


T17. Triage and validate an inbound vulnerability report

Assess whether a reported vulnerability is valid, in scope, and novel. Determine severity for internal prioritization. First step in the producer-side reactive workflow.

Role type Roles
Primary 2, 4
Secondary 34, 35

T18. Coordinate disclosure timeline, embargo, and downstream notification

Negotiate and manage the disclosure timeline with the reporting researcher, coordinate embargo among notified parties, and notify downstream OEM partners and distributors before public disclosure.

Role type Roles
Primary 2
Secondary 34, 37

T19. Reserve or assign a CVE ID

Request, reserve, or assign a CVE ID for a confirmed vulnerability. May be performed by the vendor CNA, a coordinating CNA, or a researcher working through a CERT.

Role type Roles
Primary 2, 3, 34
Secondary 22, 24

Phase 7: Record Enrichment


T20. Add or correct structured fields in an existing CVE record

Supplement a CNA-published record with missing or incorrect structured data: CVSS vectors, CWE assignments, CPE strings, package URLs, KEV status, EPSS scores, or other third-party signals. The primary ADP function.

Role type Roles
Primary 27
Secondary 34, 35, 37

Phase 8: Risk Assessment


T21. Assess the risk an unpublished vulnerability poses to inform response decisions

Estimate the risk of a vulnerability that has not yet been publicly disclosed. The practitioner owns all information that exists about the vulnerability — no external signals, exploitation telemetry, or third-party characterization are available. Risk estimation is grounded entirely in internal characterization: what the fault is, what conditions allow exploitation, and what the consequence would be. Performed under information scarcity and often under disclosure deadline pressure. The quality of this assessment depends entirely on the depth and accuracy of the practitioner’s own analytical work.

Role type Roles
Primary 1, 2, 4
Secondary 34

T22. Assess the risk a published vulnerability poses to inform response decisions

Estimate the risk of a vulnerability that is publicly known. The practitioner is a consumer of externally produced information they did not generate. Risk estimation draws on the published record, enrichment data, exploitation telemetry, threat actor targeting signals, and other ecosystem-produced intelligence. Information is richer but noisier than in T21 — conflicts between sources, stale data, and shallow characterization are the primary failure modes. The quality of this assessment depends on the accuracy, completeness, and currency of external data sources.

Role type Roles
Primary 9, 11, 12, 16, 37
Secondary 17, 38

Phase 9: Prioritization


T23. Prioritize vulnerabilities for remediation

Rank open vulnerabilities by urgency and allocate remediation resources. Combines severity, exploitability, asset criticality, network exposure, and organizational context. Drives the remediation queue.

Role type Roles
Primary 9, 10, 16
Secondary 12, 17, 37

Phase 10: Detection Engineering


T24. Develop detection signatures and rules from CVE exploitation conditions

Write network detection signatures (Snort, Suricata), endpoint detection rules (YARA, EDR behavioral), or WAF rules based on the exploitation conditions and attack behavior described in or derivable from a CVE record. Requires exploitation conditions at high specificity.

Role type Roles
Primary 13
Secondary 25, 37

Phase 11: Offensive Operations


T25. Develop or adapt a working exploit or proof-of-concept

Produce functional exploit or PoC code from CVE characterization data and references. Used for vulnerability validation, red team operations, or public demonstration. Requires fault location and exploitation conditions at high specificity.

Role type Roles
Primary 15, 23
Secondary 13, 22

Phase 12: Remediation


T26. Identify the available fix and source the patch

Determine what version resolves the vulnerability, locate the patch or update package, and confirm the fix is applicable to the affected asset. First step in fix-based remediation.

Role type Roles
Primary 9, 10
Secondary 5, 6, 14, 18

T27. Deploy patch or update to affected assets

Apply the available fix across affected systems. Includes scheduling, deployment execution, and status tracking. The physical remediation act.

Role type Roles
Primary 18
Secondary 10, 14

T28. Design and implement a workaround or compensating control

Produce and apply a control that removes exploitation conditions or reduces security outcome without eliminating the underlying weakness. Used when a patch is unavailable, untested, or cannot be deployed immediately.

Role type Roles
Primary 13, 18
Secondary 4, 6, 9

T29. Verify that a vulnerability has been successfully remediated

Confirm that a patched or mitigated asset no longer registers as vulnerable. May use scanner re-run, version confirmation, or configuration validation. Closes the remediation loop.

Role type Roles
Primary 9, 10, 18
Secondary 14, 17

Phase 13: Communication and Advisory


T30. Write a technical or customer-facing advisory or security bulletin

Produce a document communicating a vulnerability’s existence, severity, affected scope, and remediation steps to an external or internal audience. May be a formal advisory, a security bulletin, a national alert, or a customer notification.

Role type Roles
Primary 3, 6, 37
Secondary 2, 7, 22

T31. Communicate vulnerability impact to non-technical audiences

Translate CVE-based technical findings into language appropriate for executives, board members, customers, media, or regulators. Requires stripping technical detail while preserving decision-relevant content.

Role type Roles
Primary 6, 7, 16, 41
Secondary 8, 38, 40

Phase 14: Threat Intelligence Correlation


T32. Correlate CVE records with threat actor activity and exploitation telemetry

Join CVE data to threat actor campaign information, observed exploitation events, and dark web signals. Answers which CVEs are being actively weaponized, by whom, and at what scale.

Role type Roles
Primary 12, 32, 33
Secondary 9, 37

T33. Map CVE records to ATT&CK techniques and threat intelligence frameworks

Establish structured links between CVE IDs and MITRE ATT&CK techniques, CAPEC patterns, and other threat classification frameworks. Enables threat-model-driven prioritization and detection coverage analysis.

Role type Roles
Primary 33
Secondary 12, 37

Phase 15: Compliance and Regulatory Tracking


T34. Map CVEs to regulatory requirements and compliance frameworks

Determine which open CVEs are relevant to applicable regulations (PCI-DSS, FedRAMP, HIPAA, NIS2, CISA KEV mandate). Establishes which vulnerabilities carry mandatory remediation timelines.

Role type Roles
Primary 17, 36, 38
Secondary 34, 37

T35. Track CVE remediation against compliance deadlines and produce audit evidence

Monitor remediation status for compliance-relevant CVEs, escalate overdue items, and document completed remediation with timestamps and method for audit purposes.

Role type Roles
Primary 17
Secondary 9, 10

Phase 16: Vendor and Procurement Risk


T36. Assess vendor security posture for procurement or contractual decisions

Evaluate a vendor’s CVE history — volume, severity distribution, time-to-fix, advisory quality, and disclosure responsiveness — as a signal of security maturity before or during a commercial relationship.

Role type Roles
Primary 21
Secondary 8, 17

Phase 17: Data Quality Management


T37. Monitor and measure CVE data quality at corpus scale

Systematically evaluate the quality of CVE records across the corpus — field population rates, CVSS consistency, CWE validity, version evaluability — and produce metrics that identify systemic failures by CNA, product category, or time period.

Role type Roles
Primary 27, 35, 39
Secondary 30, 34, 40

T38. Resolve conflicts between sources with different characterizations

Determine the authoritative characterization when CVE, NVD, GHSA, vendor advisory, and researcher write-up disagree on CVSS, affected scope, CWE, or fix availability. Requires source weighting and provenance tracking.

Role type Roles
Primary 28, 31, 32
Secondary 27, 35

T39. Maintain provenance records for data elements across sources

Track which source made which assertion about a vulnerability, when, and on what basis. Required for conflict resolution, trust assessment, and accountability when enriched data is found to be incorrect.

Role type Roles
Primary 27, 31, 32
Secondary 28, 30

Phase 18: Analysis and Reporting


T40. Generate trend reports and aggregate analyses from CVE corpus data

Produce summary reports on vulnerability patterns: counts by product, vendor, weakness type, severity, or disclosure lag. Supports organizational risk reporting, policy development, and public communication.

Role type Roles
Primary 39, 40
Secondary 34, 37, 38, 41

T41. Measure vulnerability program effectiveness

Calculate metrics that assess how well a vulnerability management program is performing: mean time to remediate (MTTR), patch coverage rates, SLA compliance, critical vulnerability exposure windows.

Role type Roles
Primary 9, 16, 40
Secondary 34, 38, 39

T42. Conduct empirical research on the CVE ecosystem

Study the CVE ecosystem as a subject of inquiry: scoring reliability, disclosure dynamics, exploitation patterns, CNA behavior, data quality characteristics. Produces publishable findings that inform program improvement.

Role type Roles
Primary 39
Secondary 36, 40

Phase 19: Governance and Standards


T43. Define and enforce CVE program policies and CNA quality criteria

Establish and apply the rules that govern what counts as a valid CVE record, how CNAs must behave, and what quality standards apply across the program.

Role type Roles
Primary 34
Secondary 35, 36

T44. Participate in standards governance

Contribute to the development, revision, and maintenance of the controlled vocabularies, scoring rubrics, and schema specifications that CVE records reference: CVSS, CWE, CVE JSON schema.

Role type Roles
Primary 36
Secondary 27, 34, 39

T45. Define regulatory mandates using CVE-derived metrics

Establish legislation, executive orders, or regulatory requirements that reference CVE data — severity thresholds for mandatory patching timelines, KEV membership as a trigger, or CVE-based disclosure obligations.

Role type Roles
Primary 38
Secondary 37, 40

Role-to-Task Index

Lookup table: for each role, the canonical tasks they perform.

Role Primary tasks Secondary tasks
1. Internal vulnerability finder T11, T12, T13, T14, T21 T9
2. PSIRT coordinator T17, T18, T19, T21 T15, T16
3. CNA record author T14, T15, T30 T16, T19
4. Security engineer (producer) T9, T12, T13, T14, T17, T21 T28
5. Developer / fix author T26 (fix details)
6. Advisory author / technical writer T28, T30, T31 T26
7. Customer support representative T31 T10, T30
8. Sales / account manager T10, T31, T36
9. Vulnerability analyst T07, T08, T09, T10, T22, T23, T26, T29, T41
10. Patch manager T23, T26, T27, T29 T35
11. SOC analyst T10, T22
12. Threat intelligence analyst T22, T32 T23, T33
13. Detection engineer T13, T24, T28 T25
14. Application security engineer T07, T09 T08, T26, T27, T29
15. Penetration tester / red teamer T10, T25 T09, T11, T12, T13
16. Risk officer / CISO T22, T23, T31, T41 T38
17. Compliance officer T34, T35 T22, T29, T36
18. IT operations / system administrator T27, T28, T29 T26
19. Vulnerability data engineer T01, T02, T03, T04 T07 (pipeline)
20. Asset inventory / CMDB owner — (feeds T07) T07 (enabler)
21. Procurement / vendor risk analyst T36 T10
22. Independent researcher / finder T10, T11, T12, T13, T14, T30 T19, T25
23. Exploit developer T12, T25 T10
24. Bug bounty participant T11 T10, T19
25. Security tool vendor / developer T01, T02, T07, T08 T03, T24
26. SBOM tooling developer T07 (purl-first)
27. Vulnerability enricher / ADP operator T16, T20, T37, T39 T05, T12, T38
28. Vulnerability database curator T05, T06, T38 T04, T39
29. Vulnerability database operator T01, T03 T02
30. Data pipeline engineer T01, T02, T03, T04 T05, T39
31. Disambiguation / identity resolution analyst T05, T06, T38, T39 T04
32. Vulnerability feed aggregator / broker T01, T02, T04, T05, T32, T38, T39 T06
33. TIP integrator T32, T33
34. CVE program administrator / Root CNA T06, T19, T43 T15, T16, T18, T20, T21, T34, T37, T40, T41
35. CNA quality reviewer T37 T16, T17, T20, T38
36. Standards body representative T34, T44 T42, T43
37. Government / national CERT analyst T10, T22, T23, T30, T37 T18, T20, T32, T33, T34, T40
38. Policy maker / regulator T22, T34, T45 T31, T40
39. Academic / industry researcher T37, T42 T40, T41, T44
40. Data analyst T40, T41 T31, T37
41. Journalist T10, T31 T40

Task Coverage Summary

Tasks performed by the most roles (highest normalization value):

Task Primary role count Total roles
T10 Look up and interpret a CVE record 8 16
T07 Match CVE records to assets 4 7
T09 Confirm match applicability 3 5
T23 Prioritize vulnerabilities 3 6
T29 Verify remediation 3 5
T37 Monitor and measure data quality 3 6
T38 Resolve cross-source conflicts 3 5

Tasks performed by the fewest roles (highest specialization):

Task Primary roles
T15 Author and publish a CVE record 3 (CNA author only as primary)
T18 Coordinate disclosure and embargo 2 (PSIRT coordinator)
T24 Develop detection signatures 13
T33 Map CVEs to ATT&CK 33
T35 Track compliance deadlines 17
T42 Conduct empirical research 39
T43 Define program policies 34
T45 Define regulatory mandates 38

Tasks with the highest data quality sensitivity (tasks where CVE record quality failures directly produce wrong outputs):

Task Primary failure mode from poor CVE data
T07 Asset matching False positives and false negatives from identifier inconsistency
T08 Resolve false positives Entire task exists because of CVE data quality failures
T09 Confirm match applicability Cannot confirm when exploitation conditions are absent from record
T10 Look up and interpret Shallow characterization produces incomplete or incorrect understanding
T13 Enumerate exploitation conditions Record provides coarse CVSS labels, not full condition set
T24 Develop detection signatures Cannot write specific signature when attack vector is underspecified
T25 Develop exploit / PoC Cannot reproduce without fault location and condition specificity
T26 Identify fix and source patch Cannot identify fix when fixed version is not encoded or solutions field absent
T28 Design workaround Cannot design effective control when exploitation conditions are absent
T32 Correlate with threat intel Temporal misalignment from unreliable datePublic undermines correlation
T38 Resolve cross-source conflicts Conflict exists because CVE record and enrichment source disagree

Notes for Working Group

On the 45-task count. This can be compressed further. T01–T04 (pipeline tasks) are candidates for consolidation into a single “data management” cluster for the DQAF’s purposes if the framework’s assessment unit is the organizational program rather than the individual pipeline step. Retaining them separately is correct if the framework will include tool vendors and data ecosystem roles.

On data quality requirements. The next analytical step is identifying, for each canonical task, the specific CVE record fields and data properties required for the task to succeed. This produces the data requirements matrix — the analytical heart of the DQAF. The high-sensitivity tasks above are the starting point.

On the IMF DQAF parallel. The IMF DQAF’s five quality dimensions (integrity, methodological soundness, accuracy, serviceability, accessibility) were derived from the tasks national statistical offices perform in producing and disseminating national accounts data. The analogous move here is: from these 45 practitioner tasks, derive the dimensions and indicators that the CVE DQAF will use to assess record quality. The Wang and Strong and Wand and Wang frameworks supply the dimension vocabulary; the task list grounds which dimensions matter and for whom.