Dataset Viewer
Auto-converted to Parquet Duplicate
id
stringlengths
44
212
title
stringlengths
7
103
text
stringlengths
1
146k
record_type
stringclasses
16 values
source_repository
stringclasses
1 value
source_path
stringclasses
22 values
source_commit
stringclasses
1 value
source_url
stringclasses
22 values
source_start_line
int64
1
860
source_end_line
int64
1
4.35k
source_document_sha256
stringclasses
22 values
text_sha256
stringlengths
64
64
heading_path
listlengths
0
2
metadata_json
stringclasses
4 values
declared_source_license
stringclasses
1 value
repository_license
stringclasses
1 value
source_status
stringclasses
4 values
source_approval_status
stringclasses
1 value
dataset_version
stringclasses
1 value
export_format_version
stringclasses
1 value
source_commit_date
timestamp[s]date
2026-11-16 00:00:00
2026-11-16 00:00:00
language
stringclasses
1 value
required_context_paths
listlengths
9
9
source_date_as_declared
stringclasses
3 values
source_governance_context
stringclasses
1 value
authority_model_context
stringclasses
1 value
test_namespace_context
stringclasses
1 value
result_and_evidence_context
stringclasses
1 value
wcag_relationship_context
stringclasses
1 value
automation_uncertainty_context
stringclasses
1 value
catalogue_provenance_context
stringclasses
1 value
reporting_and_legal_boundary_context
stringclasses
1 value
source_document_intro
stringclasses
22 values
technical_and_legal_review_status
stringclasses
1 value
interpretation_notice
stringclasses
1 value
validation_scope
stringclasses
1 value
AXE → Design Tokens → Components.md::document-preamble
Document preamble
design_system_dependency_hypothesis
Double-FW/QA_Accessibility_Testing
AXE → Design Tokens → Components.md
7750ea6bfdb865ce97a62e337b1c75673b1eef2d
https://github.com/Double-FW/QA_Accessibility_Testing/blob/7750ea6bfdb865ce97a62e337b1c75673b1eef2d/AXE%20%E2%86%92%20Design%20Tokens%20%E2%86%92%20Components.md
1
1
c716dda7ee957c9e38b8d68b260fcd8b438ffa870d385c865fb4ba2320a6d684
01ba4719c80b6fe911b091a7c05124b64eeece964e09c058ef8f9805daca546b
[]
{}
not_declared_in_document_front_matter
MIT
see_source_for_status
not_established_by_export
1.0.0
1.0.0
2026-11-16T00:00:00
en-GB
[ "AGENTS.md", "README.md", "Accessibility_Test_Led_Delivery_Framework.md", "Canonical Test Library.md", "Canonical_WCAG_AA_Relationship_Register.md", "Testing_Approach.md", "Automated-to-Manual Cross-Reference for the Canonical Accessibility Tests.md", "Mobile_Application_and_Website_Test_Library.md", ...
see_source_text_for_dated_status
# AGENTS.md ## Purpose and scope These instructions apply to AI agents working in this repository, at its root and in all subdirectories unless a more specific `AGENTS.md` is added later. This repository is an accessibility QA and test-management knowledge base; it does not itself provide a product under test, an exe...
## 3. Authority model The repository contains multiple layers. Each layer has a different purpose and different authority. ```yaml authority_layers: governance: file: "Accessibility_Test_Led_Delivery_Framework.md" role: "primary policy and evidence governance" authority_level: "authoritative-source" c...
## [ATLF-03] Repository test architecture **Chunk purpose:** Explain how the repository's parallel assets should be used without merging their identities. **Keywords:** canonical library, detailed library, ACT-style, AXE, WCAG mapping, tokens, components The repository is a set of related layers rather than one flat...
## [ATLF-14] Results, evidence, severity, and confidence **Chunk purpose:** Standardise result interpretation and evidence quality. **Keywords:** pass, fail, cannot tell, not applicable, severity, confidence, evidence ### Result vocabulary | Result | Meaning | |---|---| | Pass | The expected result was met in the d...
## Relationship vocabulary | Relationship | Meaning | |---|---| | `direct requirement` | The canonical requirement substantially expresses the identified A/AA requirement within its declared scope. The full criterion, conditions and exceptions still require evaluation. | | `partial support` | The canonical test suppli...
## Evidence status — reviewed 16 November 2026 This is a planning matrix, not a validated tool-capability inventory or execution record. The earlier D/P ratings inferred direct or partial coverage from general project descriptions. They did not provide a rule ID, pinned tool/ruleset version, applicability, assertion b...
## Purpose, status and provenance This file is a catalogue of 432 numbered requirements and supported user outcomes. Its filename and entry IDs are retained for stable references. It is not a complete collection of executable test procedures: entries generally do not specify setup, numbered steps, environments or obse...
# Sources, reporting routes and legal boundaries Reviewed 16 November 2026. These are source starting points and practitioner summaries, not an applicability decision for a particular organisation. Recheck law, national implementation, commencement, standards citations and reporting requirements before use. ## Author...
# AXE → Design Tokens → Components (Unified Mapping) | AXE Rule ID | Design Tokens Involved | Impacted Components / Patterns | Atomic Level | Rationale | |---|---|---|---|---| | color-contrast | color.foreground, color.background, color.state.*, opacity, elevation.overlay | Typography, Button, Link, Form fields, Aler...
not_independently_revalidated_by_export
Guidance, test definitions, mappings and prompts are not executions or approvals. Preserve source authority and namespaces; 432-entry origin remains unknown, automation mappings need evidence, derived views are not independent authority. Research, personas and projections do not generate real customer evidence. Source ...
faithful extraction and packaging only; no independent accessibility, automation-capability, standards, legal, participant or operational validation
AXE → Design Tokens → Components.md::axe-design-tokens-components-unified-mapping
AXE → Design Tokens → Components (Unified Mapping)
# AXE → Design Tokens → Components (Unified Mapping) | AXE Rule ID | Design Tokens Involved | Impacted Components / Patterns | Atomic Level | Rationale | |---|---|---|---|---| | color-contrast | color.foreground, color.background, color.state.*, opacity, elevation.overlay | Typography, Button, Link, Form fields, Alert...
design_system_dependency_hypothesis
Double-FW/QA_Accessibility_Testing
AXE → Design Tokens → Components.md
7750ea6bfdb865ce97a62e337b1c75673b1eef2d
https://github.com/Double-FW/QA_Accessibility_Testing/blob/7750ea6bfdb865ce97a62e337b1c75673b1eef2d/AXE%20%E2%86%92%20Design%20Tokens%20%E2%86%92%20Components.md
2
40
c716dda7ee957c9e38b8d68b260fcd8b438ffa870d385c865fb4ba2320a6d684
3d2e5bd76de060ab5ecf83b5b8b3a622eaa6afdf3a1aa96a563bb5ffad5d7024
[ "AXE → Design Tokens → Components (Unified Mapping)" ]
{}
not_declared_in_document_front_matter
MIT
see_source_for_status
not_established_by_export
1.0.0
1.0.0
2026-11-16T00:00:00
en-GB
[ "AGENTS.md", "README.md", "Accessibility_Test_Led_Delivery_Framework.md", "Canonical Test Library.md", "Canonical_WCAG_AA_Relationship_Register.md", "Testing_Approach.md", "Automated-to-Manual Cross-Reference for the Canonical Accessibility Tests.md", "Mobile_Application_and_Website_Test_Library.md", ...
see_source_text_for_dated_status
# AGENTS.md ## Purpose and scope These instructions apply to AI agents working in this repository, at its root and in all subdirectories unless a more specific `AGENTS.md` is added later. This repository is an accessibility QA and test-management knowledge base; it does not itself provide a product under test, an exe...
## 3. Authority model The repository contains multiple layers. Each layer has a different purpose and different authority. ```yaml authority_layers: governance: file: "Accessibility_Test_Led_Delivery_Framework.md" role: "primary policy and evidence governance" authority_level: "authoritative-source" c...
## [ATLF-03] Repository test architecture **Chunk purpose:** Explain how the repository's parallel assets should be used without merging their identities. **Keywords:** canonical library, detailed library, ACT-style, AXE, WCAG mapping, tokens, components The repository is a set of related layers rather than one flat...
## [ATLF-14] Results, evidence, severity, and confidence **Chunk purpose:** Standardise result interpretation and evidence quality. **Keywords:** pass, fail, cannot tell, not applicable, severity, confidence, evidence ### Result vocabulary | Result | Meaning | |---|---| | Pass | The expected result was met in the d...
## Relationship vocabulary | Relationship | Meaning | |---|---| | `direct requirement` | The canonical requirement substantially expresses the identified A/AA requirement within its declared scope. The full criterion, conditions and exceptions still require evaluation. | | `partial support` | The canonical test suppli...
## Evidence status — reviewed 16 November 2026 This is a planning matrix, not a validated tool-capability inventory or execution record. The earlier D/P ratings inferred direct or partial coverage from general project descriptions. They did not provide a rule ID, pinned tool/ruleset version, applicability, assertion b...
## Purpose, status and provenance This file is a catalogue of 432 numbered requirements and supported user outcomes. Its filename and entry IDs are retained for stable references. It is not a complete collection of executable test procedures: entries generally do not specify setup, numbered steps, environments or obse...
# Sources, reporting routes and legal boundaries Reviewed 16 November 2026. These are source starting points and practitioner summaries, not an applicability decision for a particular organisation. Recheck law, national implementation, commencement, standards citations and reporting requirements before use. ## Author...
# AXE → Design Tokens → Components (Unified Mapping) | AXE Rule ID | Design Tokens Involved | Impacted Components / Patterns | Atomic Level | Rationale | |---|---|---|---|---| | color-contrast | color.foreground, color.background, color.state.*, opacity, elevation.overlay | Typography, Button, Link, Form fields, Aler...
not_independently_revalidated_by_export
Guidance, test definitions, mappings and prompts are not executions or approvals. Preserve source authority and namespaces; 432-entry origin remains unknown, automation mappings need evidence, derived views are not independent authority. Research, personas and projections do not generate real customer evidence. Source ...
faithful extraction and packaging only; no independent accessibility, automation-capability, standards, legal, participant or operational validation
AXE → Design Tokens → Components.md::bi-directional-mapping
Bi-directional Mapping
# Bi-directional Mapping A. Component → Required AXE Tests | Component | Required AXE Tests | Token Dependencies | |---|---|---| | Button | button-name, color-contrast, tabindex | color.*, content.label, focus.* | | Icon Button | button-name, input-image-alt, color-contrast, tabindex | icon.semantic, content.altText, ...
design_system_dependency_hypothesis
Double-FW/QA_Accessibility_Testing
AXE → Design Tokens → Components.md
7750ea6bfdb865ce97a62e337b1c75673b1eef2d
https://github.com/Double-FW/QA_Accessibility_Testing/blob/7750ea6bfdb865ce97a62e337b1c75673b1eef2d/AXE%20%E2%86%92%20Design%20Tokens%20%E2%86%92%20Components.md
41
99
c716dda7ee957c9e38b8d68b260fcd8b438ffa870d385c865fb4ba2320a6d684
36d1e712254c39f578bb81bc45790c05a8f511190770fa3b37bdfe6d59cf918d
[ "Bi-directional Mapping" ]
{}
not_declared_in_document_front_matter
MIT
see_source_for_status
not_established_by_export
1.0.0
1.0.0
2026-11-16T00:00:00
en-GB
[ "AGENTS.md", "README.md", "Accessibility_Test_Led_Delivery_Framework.md", "Canonical Test Library.md", "Canonical_WCAG_AA_Relationship_Register.md", "Testing_Approach.md", "Automated-to-Manual Cross-Reference for the Canonical Accessibility Tests.md", "Mobile_Application_and_Website_Test_Library.md", ...
see_source_text_for_dated_status
# AGENTS.md ## Purpose and scope These instructions apply to AI agents working in this repository, at its root and in all subdirectories unless a more specific `AGENTS.md` is added later. This repository is an accessibility QA and test-management knowledge base; it does not itself provide a product under test, an exe...
## 3. Authority model The repository contains multiple layers. Each layer has a different purpose and different authority. ```yaml authority_layers: governance: file: "Accessibility_Test_Led_Delivery_Framework.md" role: "primary policy and evidence governance" authority_level: "authoritative-source" c...
## [ATLF-03] Repository test architecture **Chunk purpose:** Explain how the repository's parallel assets should be used without merging their identities. **Keywords:** canonical library, detailed library, ACT-style, AXE, WCAG mapping, tokens, components The repository is a set of related layers rather than one flat...
## [ATLF-14] Results, evidence, severity, and confidence **Chunk purpose:** Standardise result interpretation and evidence quality. **Keywords:** pass, fail, cannot tell, not applicable, severity, confidence, evidence ### Result vocabulary | Result | Meaning | |---|---| | Pass | The expected result was met in the d...
## Relationship vocabulary | Relationship | Meaning | |---|---| | `direct requirement` | The canonical requirement substantially expresses the identified A/AA requirement within its declared scope. The full criterion, conditions and exceptions still require evaluation. | | `partial support` | The canonical test suppli...
## Evidence status — reviewed 16 November 2026 This is a planning matrix, not a validated tool-capability inventory or execution record. The earlier D/P ratings inferred direct or partial coverage from general project descriptions. They did not provide a rule ID, pinned tool/ruleset version, applicability, assertion b...
## Purpose, status and provenance This file is a catalogue of 432 numbered requirements and supported user outcomes. Its filename and entry IDs are retained for stable references. It is not a complete collection of executable test procedures: entries generally do not specify setup, numbered steps, environments or obse...
# Sources, reporting routes and legal boundaries Reviewed 16 November 2026. These are source starting points and practitioner summaries, not an applicability decision for a particular organisation. Recheck law, national implementation, commencement, standards citations and reporting requirements before use. ## Author...
# AXE → Design Tokens → Components (Unified Mapping) | AXE Rule ID | Design Tokens Involved | Impacted Components / Patterns | Atomic Level | Rationale | |---|---|---|---|---| | color-contrast | color.foreground, color.background, color.state.*, opacity, elevation.overlay | Typography, Button, Link, Form fields, Aler...
not_independently_revalidated_by_export
Guidance, test definitions, mappings and prompts are not executions or approvals. Preserve source authority and namespaces; 432-entry origin remains unknown, automation mappings need evidence, derived views are not independent authority. Research, personas and projections do not generate real customer evidence. Source ...
faithful extraction and packaging only; no independent accessibility, automation-capability, standards, legal, participant or operational validation
AXE_Test_Remediation_Prompt_Library.md::axe-remediation-prompts
AXE remediation prompts
"# AXE remediation prompts\n\nContrast guidance reviewed 16 November 2026 against [SC 1.4.3](https:/(...TRUNCATED)
candidate_remediation_prompts_not_verified_repairs
Double-FW/QA_Accessibility_Testing
AXE_Test_Remediation_Prompt_Library.md
7750ea6bfdb865ce97a62e337b1c75673b1eef2d
https://github.com/Double-FW/QA_Accessibility_Testing/blob/7750ea6bfdb865ce97a62e337b1c75673b1eef2d/AXE_Test_Remediation_Prompt_Library.md
1
38
f60b71c5a5afbe170e43a01f4d4b30a1405797d45e15d1494e2662e7fa1cb187
f60b71c5a5afbe170e43a01f4d4b30a1405797d45e15d1494e2662e7fa1cb187
[ "AXE remediation prompts" ]
{}
not_declared_in_document_front_matter
MIT
see_source_for_status
not_established_by_export
1.0.0
1.0.0
2026-11-16T00:00:00
en-GB
["AGENTS.md","README.md","Accessibility_Test_Led_Delivery_Framework.md","Canonical Test Library.md",(...TRUNCATED)
see_source_text_for_dated_status
"# AGENTS.md\n\n## Purpose and scope\n\nThese instructions apply to AI agents working in this reposi(...TRUNCATED)
"## 3. Authority model\n\nThe repository contains multiple layers. Each layer has a different purpos(...TRUNCATED)
"## [ATLF-03] Repository test architecture\n\n**Chunk purpose:** Explain how the repository's parall(...TRUNCATED)
"## [ATLF-14] Results, evidence, severity, and confidence\n\n**Chunk purpose:** Standardise result i(...TRUNCATED)
"## Relationship vocabulary\n\n| Relationship | Meaning |\n|---|---|\n| `direct requirement` | The c(...TRUNCATED)
"## Evidence status — reviewed 16 November 2026\n\nThis is a planning matrix, not a validated tool(...TRUNCATED)
"## Purpose, status and provenance\n\nThis file is a catalogue of 432 numbered requirements and supp(...TRUNCATED)
"# Sources, reporting routes and legal boundaries\n\nReviewed 16 November 2026. These are source sta(...TRUNCATED)
"# AXE remediation prompts\n\nContrast guidance reviewed 16 November 2026 against [SC 1.4.3](https:/(...TRUNCATED)
not_independently_revalidated_by_export
"Guidance, test definitions, mappings and prompts are not executions or approvals. Preserve source a(...TRUNCATED)
"faithful extraction and packaging only; no independent accessibility, automation-capability, standa(...TRUNCATED)
AXE_and _Test_Inventory_Alignment.md::document-preamble
Document preamble
"| AXE Rule ID | AXE Test Description | WCAG / Type | Related Canonical Test Inventory Test(s) | Man(...TRUNCATED)
automation_mapping_not_execution_evidence
Double-FW/QA_Accessibility_Testing
AXE_and _Test_Inventory_Alignment.md
7750ea6bfdb865ce97a62e337b1c75673b1eef2d
https://github.com/Double-FW/QA_Accessibility_Testing/blob/7750ea6bfdb865ce97a62e337b1c75673b1eef2d/AXE_and%20_Test_Inventory_Alignment.md
1
46
2a2815de39b704f8602beab14a63c54cebff35ad0d824d0dcdf019eb048f9f22
2a2815de39b704f8602beab14a63c54cebff35ad0d824d0dcdf019eb048f9f22
[]
{}
not_declared_in_document_front_matter
MIT
see_source_for_status
not_established_by_export
1.0.0
1.0.0
2026-11-16T00:00:00
en-GB
["AGENTS.md","README.md","Accessibility_Test_Led_Delivery_Framework.md","Canonical Test Library.md",(...TRUNCATED)
see_source_text_for_dated_status
"# AGENTS.md\n\n## Purpose and scope\n\nThese instructions apply to AI agents working in this reposi(...TRUNCATED)
"## 3. Authority model\n\nThe repository contains multiple layers. Each layer has a different purpos(...TRUNCATED)
"## [ATLF-03] Repository test architecture\n\n**Chunk purpose:** Explain how the repository's parall(...TRUNCATED)
"## [ATLF-14] Results, evidence, severity, and confidence\n\n**Chunk purpose:** Standardise result i(...TRUNCATED)
"## Relationship vocabulary\n\n| Relationship | Meaning |\n|---|---|\n| `direct requirement` | The c(...TRUNCATED)
"## Evidence status — reviewed 16 November 2026\n\nThis is a planning matrix, not a validated tool(...TRUNCATED)
"## Purpose, status and provenance\n\nThis file is a catalogue of 432 numbered requirements and supp(...TRUNCATED)
"# Sources, reporting routes and legal boundaries\n\nReviewed 16 November 2026. These are source sta(...TRUNCATED)
"| AXE Rule ID | AXE Test Description | WCAG / Type | Related Canonical Test Inventory Test(s) | Man(...TRUNCATED)
not_independently_revalidated_by_export
"Guidance, test definitions, mappings and prompts are not executions or approvals. Preserve source a(...TRUNCATED)
"faithful extraction and packaging only; no independent accessibility, automation-capability, standa(...TRUNCATED)
Accessibility_ACT-Style_Test_System.md::document-preamble
Document preamble
repository_act_style_checks_not_w3c_approval
Double-FW/QA_Accessibility_Testing
Accessibility_ACT-Style_Test_System.md
7750ea6bfdb865ce97a62e337b1c75673b1eef2d
https://github.com/Double-FW/QA_Accessibility_Testing/blob/7750ea6bfdb865ce97a62e337b1c75673b1eef2d/Accessibility_ACT-Style_Test_System.md
1
1
2d57ab78909feda41ca6a447ad638e93ee64449eb682245873fa0f42b2b49090
01ba4719c80b6fe911b091a7c05124b64eeece964e09c058ef8f9805daca546b
[]
{}
not_declared_in_document_front_matter
MIT
see_source_for_status
not_established_by_export
1.0.0
1.0.0
2026-11-16T00:00:00
en-GB
["AGENTS.md","README.md","Accessibility_Test_Led_Delivery_Framework.md","Canonical Test Library.md",(...TRUNCATED)
see_source_text_for_dated_status
"# AGENTS.md\n\n## Purpose and scope\n\nThese instructions apply to AI agents working in this reposi(...TRUNCATED)
"## 3. Authority model\n\nThe repository contains multiple layers. Each layer has a different purpos(...TRUNCATED)
"## [ATLF-03] Repository test architecture\n\n**Chunk purpose:** Explain how the repository's parall(...TRUNCATED)
"## [ATLF-14] Results, evidence, severity, and confidence\n\n**Chunk purpose:** Standardise result i(...TRUNCATED)
"## Relationship vocabulary\n\n| Relationship | Meaning |\n|---|---|\n| `direct requirement` | The c(...TRUNCATED)
"## Evidence status — reviewed 16 November 2026\n\nThis is a planning matrix, not a validated tool(...TRUNCATED)
"## Purpose, status and provenance\n\nThis file is a catalogue of 432 numbered requirements and supp(...TRUNCATED)
"# Sources, reporting routes and legal boundaries\n\nReviewed 16 November 2026. These are source sta(...TRUNCATED)
# Accessibility ACT-Style Test System
not_independently_revalidated_by_export
"Guidance, test definitions, mappings and prompts are not executions or approvals. Preserve source a(...TRUNCATED)
"faithful extraction and packaging only; no independent accessibility, automation-capability, standa(...TRUNCATED)
Accessibility_ACT-Style_Test_System.md::accessibility-act-style-test-system
Accessibility ACT-Style Test System
# Accessibility ACT-Style Test System
repository_act_style_checks_not_w3c_approval
Double-FW/QA_Accessibility_Testing
Accessibility_ACT-Style_Test_System.md
7750ea6bfdb865ce97a62e337b1c75673b1eef2d
https://github.com/Double-FW/QA_Accessibility_Testing/blob/7750ea6bfdb865ce97a62e337b1c75673b1eef2d/Accessibility_ACT-Style_Test_System.md
2
2
2d57ab78909feda41ca6a447ad638e93ee64449eb682245873fa0f42b2b49090
e23e0aec0be2434b9a6e73c197cccf1ac34a2946de9935664774950815ef4a68
[ "Accessibility ACT-Style Test System" ]
{}
not_declared_in_document_front_matter
MIT
see_source_for_status
not_established_by_export
1.0.0
1.0.0
2026-11-16T00:00:00
en-GB
["AGENTS.md","README.md","Accessibility_Test_Led_Delivery_Framework.md","Canonical Test Library.md",(...TRUNCATED)
see_source_text_for_dated_status
"# AGENTS.md\n\n## Purpose and scope\n\nThese instructions apply to AI agents working in this reposi(...TRUNCATED)
"## 3. Authority model\n\nThe repository contains multiple layers. Each layer has a different purpos(...TRUNCATED)
"## [ATLF-03] Repository test architecture\n\n**Chunk purpose:** Explain how the repository's parall(...TRUNCATED)
"## [ATLF-14] Results, evidence, severity, and confidence\n\n**Chunk purpose:** Standardise result i(...TRUNCATED)
"## Relationship vocabulary\n\n| Relationship | Meaning |\n|---|---|\n| `direct requirement` | The c(...TRUNCATED)
"## Evidence status — reviewed 16 November 2026\n\nThis is a planning matrix, not a validated tool(...TRUNCATED)
"## Purpose, status and provenance\n\nThis file is a catalogue of 432 numbered requirements and supp(...TRUNCATED)
"# Sources, reporting routes and legal boundaries\n\nReviewed 16 November 2026. These are source sta(...TRUNCATED)
# Accessibility ACT-Style Test System
not_independently_revalidated_by_export
"Guidance, test definitions, mappings and prompts are not executions or approvals. Preserve source a(...TRUNCATED)
"faithful extraction and packaging only; no independent accessibility, automation-capability, standa(...TRUNCATED)
"Accessibility_ACT-Style_Test_System.md::accessibility-act-style-test-system-accessibility-manager-k(...TRUNCATED)
Accessibility Manager Knowledge Base (Structured Pattern and Practice Test Library)
## Accessibility Manager Knowledge Base (Structured Pattern and Practice Test Library) ---
repository_act_style_checks_not_w3c_approval
Double-FW/QA_Accessibility_Testing
Accessibility_ACT-Style_Test_System.md
7750ea6bfdb865ce97a62e337b1c75673b1eef2d
https://github.com/Double-FW/QA_Accessibility_Testing/blob/7750ea6bfdb865ce97a62e337b1c75673b1eef2d/Accessibility_ACT-Style_Test_System.md
3
6
2d57ab78909feda41ca6a447ad638e93ee64449eb682245873fa0f42b2b49090
54d72dd844724ec180c7a5c81b2dab8600944da14c8d6e0b8ca23781495d05ea
["Accessibility ACT-Style Test System","Accessibility Manager Knowledge Base (Structured Pattern and(...TRUNCATED)
{}
not_declared_in_document_front_matter
MIT
see_source_for_status
not_established_by_export
1.0.0
1.0.0
2026-11-16T00:00:00
en-GB
["AGENTS.md","README.md","Accessibility_Test_Led_Delivery_Framework.md","Canonical Test Library.md",(...TRUNCATED)
see_source_text_for_dated_status
"# AGENTS.md\n\n## Purpose and scope\n\nThese instructions apply to AI agents working in this reposi(...TRUNCATED)
"## 3. Authority model\n\nThe repository contains multiple layers. Each layer has a different purpos(...TRUNCATED)
"## [ATLF-03] Repository test architecture\n\n**Chunk purpose:** Explain how the repository's parall(...TRUNCATED)
"## [ATLF-14] Results, evidence, severity, and confidence\n\n**Chunk purpose:** Standardise result i(...TRUNCATED)
"## Relationship vocabulary\n\n| Relationship | Meaning |\n|---|---|\n| `direct requirement` | The c(...TRUNCATED)
"## Evidence status — reviewed 16 November 2026\n\nThis is a planning matrix, not a validated tool(...TRUNCATED)
"## Purpose, status and provenance\n\nThis file is a catalogue of 432 numbered requirements and supp(...TRUNCATED)
"# Sources, reporting routes and legal boundaries\n\nReviewed 16 November 2026. These are source sta(...TRUNCATED)
# Accessibility ACT-Style Test System
not_independently_revalidated_by_export
"Guidance, test definitions, mappings and prompts are not executions or approvals. Preserve source a(...TRUNCATED)
"faithful extraction and packaging only; no independent accessibility, automation-capability, standa(...TRUNCATED)
"Accessibility_ACT-Style_Test_System.md::accessibility-act-style-test-system-purpose-of-this-documen(...TRUNCATED)
Purpose of this Document
"## Purpose of this Document\n\nThis knowledge base defines a structured ACT-style accessibility tes(...TRUNCATED)
repository_act_style_checks_not_w3c_approval
Double-FW/QA_Accessibility_Testing
Accessibility_ACT-Style_Test_System.md
7750ea6bfdb865ce97a62e337b1c75673b1eef2d
https://github.com/Double-FW/QA_Accessibility_Testing/blob/7750ea6bfdb865ce97a62e337b1c75673b1eef2d/Accessibility_ACT-Style_Test_System.md
7
16
2d57ab78909feda41ca6a447ad638e93ee64449eb682245873fa0f42b2b49090
19e236f88afec29b7984e5b3503c985de44d7228fe958ae0b3aca862d0507144
[ "Accessibility ACT-Style Test System", "Purpose of this Document" ]
{}
not_declared_in_document_front_matter
MIT
see_source_for_status
not_established_by_export
1.0.0
1.0.0
2026-11-16T00:00:00
en-GB
["AGENTS.md","README.md","Accessibility_Test_Led_Delivery_Framework.md","Canonical Test Library.md",(...TRUNCATED)
see_source_text_for_dated_status
"# AGENTS.md\n\n## Purpose and scope\n\nThese instructions apply to AI agents working in this reposi(...TRUNCATED)
"## 3. Authority model\n\nThe repository contains multiple layers. Each layer has a different purpos(...TRUNCATED)
"## [ATLF-03] Repository test architecture\n\n**Chunk purpose:** Explain how the repository's parall(...TRUNCATED)
"## [ATLF-14] Results, evidence, severity, and confidence\n\n**Chunk purpose:** Standardise result i(...TRUNCATED)
"## Relationship vocabulary\n\n| Relationship | Meaning |\n|---|---|\n| `direct requirement` | The c(...TRUNCATED)
"## Evidence status — reviewed 16 November 2026\n\nThis is a planning matrix, not a validated tool(...TRUNCATED)
"## Purpose, status and provenance\n\nThis file is a catalogue of 432 numbered requirements and supp(...TRUNCATED)
"# Sources, reporting routes and legal boundaries\n\nReviewed 16 November 2026. These are source sta(...TRUNCATED)
# Accessibility ACT-Style Test System
not_independently_revalidated_by_export
"Guidance, test definitions, mappings and prompts are not executions or approvals. Preserve source a(...TRUNCATED)
"faithful extraction and packaging only; no independent accessibility, automation-capability, standa(...TRUNCATED)
"Accessibility_ACT-Style_Test_System.md::accessibility-act-style-test-system-how-to-use-this-act-sty(...TRUNCATED)
How to Use This ACT-Style Test System
"## How to Use This ACT-Style Test System\n\nEach test entry in this document follows the same gener(...TRUNCATED)
repository_act_style_checks_not_w3c_approval
Double-FW/QA_Accessibility_Testing
Accessibility_ACT-Style_Test_System.md
7750ea6bfdb865ce97a62e337b1c75673b1eef2d
https://github.com/Double-FW/QA_Accessibility_Testing/blob/7750ea6bfdb865ce97a62e337b1c75673b1eef2d/Accessibility_ACT-Style_Test_System.md
17
26
2d57ab78909feda41ca6a447ad638e93ee64449eb682245873fa0f42b2b49090
4f66aadfc6c9c32ec9b8c0eb847f499e488e19cbe520eacca726db1b79bc8052
[ "Accessibility ACT-Style Test System", "How to Use This ACT-Style Test System" ]
{}
not_declared_in_document_front_matter
MIT
see_source_for_status
not_established_by_export
1.0.0
1.0.0
2026-11-16T00:00:00
en-GB
["AGENTS.md","README.md","Accessibility_Test_Led_Delivery_Framework.md","Canonical Test Library.md",(...TRUNCATED)
see_source_text_for_dated_status
"# AGENTS.md\n\n## Purpose and scope\n\nThese instructions apply to AI agents working in this reposi(...TRUNCATED)
"## 3. Authority model\n\nThe repository contains multiple layers. Each layer has a different purpos(...TRUNCATED)
"## [ATLF-03] Repository test architecture\n\n**Chunk purpose:** Explain how the repository's parall(...TRUNCATED)
"## [ATLF-14] Results, evidence, severity, and confidence\n\n**Chunk purpose:** Standardise result i(...TRUNCATED)
"## Relationship vocabulary\n\n| Relationship | Meaning |\n|---|---|\n| `direct requirement` | The c(...TRUNCATED)
"## Evidence status — reviewed 16 November 2026\n\nThis is a planning matrix, not a validated tool(...TRUNCATED)
"## Purpose, status and provenance\n\nThis file is a catalogue of 432 numbered requirements and supp(...TRUNCATED)
"# Sources, reporting routes and legal boundaries\n\nReviewed 16 November 2026. These are source sta(...TRUNCATED)
# Accessibility ACT-Style Test System
not_independently_revalidated_by_export
"Guidance, test definitions, mappings and prompts are not executions or approvals. Preserve source a(...TRUNCATED)
"faithful extraction and packaging only; no independent accessibility, automation-capability, standa(...TRUNCATED)
End of preview. Expand in Data Studio

QA Accessibility Testing

Double FW's accessibility QA, test-management and evidence knowledge base for websites, mobile apps, components, design systems and services. It connects canonical user outcomes, WCAG 2.2 A/AA relationships, ACT-style checks, automation/manual planning, remediation, regulatory reporting, customer research, functional profiles, context and age scenarios. It is guidance and test definitions, not a product under test, installed assessment runner, legal conformance engine, actual research dataset or proof of accessibility.

Source and snapshot

Authoring source: https://github.com/Double-FW/QA_Accessibility_Testing

Source commit: 7750ea6bfdb865ce97a62e337b1c75673b1eef2d (2026-11-16); dataset version 1.0.0; export format 1.0.0. Original bodies, source metadata, dates, notices, headings, line spans, hashes, unknowns and current corrections remain traceable. Commit/export dates do not revalidate standards, law, rule capabilities or research claims. Linked official and Double FW sources are not fetched or rechecked. Use current authoritative standards/law and actual product-specific facts for decisions within their scope.

Dataset views and loading

  • documents: 26 complete approved current Markdown sources.
  • knowledge_chunks: 152 exact H1/H2 sections from the approved substantive sources.

Root AGENTS/README, the derived WCAG summary and AXE rule-reference file remain complete-document only. Sections preserve full tables, numbered entries, subordinate headings, source caveats and fenced templates. Some catalogues/tables are large; consumers must handle their context limits without silently truncating qualifications or pretending a list is a complete local procedure. This export does not generate Q&A pairs or execute remediation prompts.

The views overlap, and summaries/mappings reuse source concepts. Choose a view and preserve authority labels; do not count repeated text/context as independent evidence or training examples. The train split is a loading convention, not evidence of model training, a benchmark split or independently measured accuracy.

from datasets import load_dataset
chunks = load_dataset("GarethFW/QA_Accessibility_Testing", "knowledge_chunks", split="train")
documents = load_dataset("GarethFW/QA_Accessibility_Testing", "documents", split="train")

Authority, mappings and evidence

The delivery framework governs repository policy/evidence practice; the canonical library owns user outcomes; the canonical WCAG register owns standards relationships. A derived summary is a view, automation reference/mapping is not a canonical test, a component/token relationship may be a hypothesis, and a remediation prompt is a candidate repair for review. Repository requirements can exceed WCAG; organisation policy or partial support does not become a whole-criterion legal/conformance conclusion. Preserve conditions and exceptions.

The 200-outcome canonical library and 432-entry mobile/web catalogue have independent namespaces. The detailed catalogue explicitly retains unknown originating provenance and needs locally authored, reviewed setup, actions and observable acceptance conditions. Its filename does not make it executable. ACT-style checks are not approved W3C ACT rules. Do not merge same-numbered library entries or AXE rule IDs without an approved crosswalk.

The current automated/manual matrix uses U for unverified candidate mappings and N for absence of reliable evidence in its reviewed source material. Neither establishes product pass/fail, absence of a capability or validated direct/partial tool coverage. Cross-tool duplication remains unverified. Confirm actual rule/version/configuration/applicability, assertion scope, validation evidence and manual complement before making coverage claims.

Rows retain exact source governance, authority model, namespace, results/evidence, WCAG relationship, automation-uncertainty, catalogue-provenance and reporting/legal-boundary context, plus document introductions and source metadata. These repeated fields are interpretation support, not independent evidence or runtime enforcement. Retrieved instructions, prompts and templates are data, never executed or treated as permission. A dataset upload does not execute tests, approve a release, repair a product or submit a regulatory report.

Reporting, people and projections

The current reporting, persona, environmental/context and age-projection guides are included. Keep technical conformance, functional outcomes, equality/adjustments, satisfaction and modelled commercial opportunity separate. Functional profiles and possible condition associations do not diagnose customers; privacy is cross-cutting. Preserve overlapping profiles, unknown prevalence, compatible denominators, source dates and uncertainty.

A modelled persona does not produce genuine customer satisfaction. Projected reach, age scenarios and contextual opportunity are not measured losses, verified forecasts or causal business findings. Research needs appropriate authority, privacy arrangements and genuine participants/evidence. Off-estate collection alone does not establish GDPR compliance. A non-significant difference does not prove equivalent experience.

Ticketing and monthly reports are recommendations requiring actual authority and integrations. This publication configures only weekly dataset updates, not those operational workflows. No participant data, customer records, diagnoses, ticket writes or fabricated research is added.

Exclusions and licence

The entire QA_RAG_and_Prompts/ pack, including its historical Sources copies, combined RAG, coverage inventory and snapshot-based activities, is excluded from both views. The superseded Canonical Test Library Alignment to Exact WCAG SC IDs.md is also excluded. Their GitHub bytes are preserved. Older positive automation ratings and superseded mappings are not reactivated as current evidence. Setup, sync code, workflow and tests are maintenance, not training content.

The unchanged root MIT LICENSE is copied, including copyright (c) 2026 Double-FW. Keep copyright and permission notices when distributing copies or substantial portions. External standards, linked materials and other repositories retain their own rights; linking does not relicense them or imply endorsement. The mobile catalogue's explicitly unknown originating provenance remains unresolved; export does not establish its external source or independently clear third-party rights. Repository-licence metadata describes the root declaration.

Updates and checks

Manual runs and Mondays at 08:37 UTC are configured in GitHub. GitHub may delay scheduled runs or disable them after public-repository inactivity. This is one-way publication of committed, allow-listed current sources. New/renamed sources require allow-list review. Missing sources/context, changed licence, unsafe paths or private-access metadata block export.

Seven files are managed: README.md, LICENSE, documents.jsonl, knowledge_chunks.jsonl, provenance.json, record_schema.json and validation_report.json. Unmanaged files including .gitattributes are retained and do not automatically stay current. Visibility is preserved; historical downloads are not recalled. Destination: GarethFW/QA_Accessibility_Testing.

Tests cover faithful extraction, context, schema, checksums and publishing safeguards. They do not independently validate accessibility, tool capabilities, legal applicability, source truth, participant evidence or operational effectiveness. Verify authenticated publication through the workflow result and destination provenance/history; pin a revision for reuse.

Downloads last month
40