Forge accessibility

forge-accessibility: Accessibility

Purpose

Audit conformance with WCAG 2.2 AA using automated evidence plus keyboard and assistive-technology reasoning.

Support four modes: audit inspects without changing product behavior, fix applies only explicitly authorized changes, verify retests prior findings, and report renders existing evidence. If no mode is supplied, use audit.

Trigger conditions

Use this module when a request names forge-accessibility, asks about accessibility, or discovery finds an applicable boundary. Run it from the repository root after project discovery.

When it applies

• Any user or operator interface
• Generated documents with interactive navigation

When it does not apply

• A protocol-only library with no human-facing output

Do not silently skip it. Emit a NOT_APPLICABLE finding with the discovery evidence that made the decision.

Inputs from project discovery

• rendered routes
• component source
• automated accessibility command

Prefer .forge/project-profile.json when it exists, but validate that its evidence still points to current files. Read ../fullstack-forge/references/PROTOCOL.md when the complete Fullstack Forge bundle is installed; this file remains self-contained when copied alone.

Inspection procedure

1. Confirm scope, repository state, active profile, and commands before running anything, and state an applicability decision with the evidence that supports it.
2. Traverse each representative screen with keyboard only: record tab order, focus visibility, traps, and any control unreachable without a pointer.
3. Compute accessible names and roles for interactive elements from markup or the accessibility tree, not from visual labels.
4. Check contrast for text and interactive states against WCAG 2.2 AA thresholds using measured colors.
5. Exercise dynamic behavior: dialogs, expanders, validation errors, and live updates, and record what assistive technology is told at each change.
6. Test zoom to 200% and reflow at 320px width, recording clipped or unusable content.
7. Run an automated scanner when available, then state plainly which criteria it cannot judge and mark those manual.
8. Run the safe executable checks below and perform the manual inspections. Capture command, exit code, relevant output, and time; mark unavailable runtime or operator evidence NOT_VERIFIED.
9. Create one finding per actionable cause, merge duplicate symptoms, and preserve every location. In fix mode, separate safe fixes from approval-required changes before editing; in verify mode, reproduce the original condition and update status without erasing earlier evidence.

Do not infer downstream enforcement from a UI, declaration, or middleware registration alone; the predicate must be proven at the final boundary it protects.

Concrete checks

• Check names, roles, states, landmarks, headings, labels, alternatives, contrast, zoom, reflow, target size, and status messages
• Complete all critical workflows using keyboard only
• Inspect focus order, focus trapping, restoration, errors, dynamic updates, and reduced motion

Required inspection criteria

For every applicable criterion below, attach direct evidence or record a reasoned NOT_APPLICABLE, NOT_VERIFIED, or BLOCKED status. The list is a routing checklist, not evidence by itself.

• WCAG 2.2 AA scope
• Keyboard navigation
• Focus order
• Focus visibility
• Focus trapping
• Semantic HTML
• Accessible names
• Form labels
• Error announcements
• Status announcements
• Contrast
• Touch targets
• Heading structure
• Landmarks
• Alt text
• Link purpose
• Table semantics
• Reduced motion
• Zoom
• Text resizing
• Responsive reflow
• Dialog behavior
• Skip links
• Screen-reader navigation
• Captions where applicable
• Auto-playing media
• Time limits
• Automated evidence plus manual keyboard and assistive-technology checks

Safe executable checks

• Run forge accessibility audit --json or fullstack-forge accessibility audit --json when the CLI is installed.
• Use inspect-rendered-ui for its bounded evidence when present; treat unavailable runtime evidence as NOT_VERIFIED.
• Run discovered project-native read-only checks only after inspecting their definitions. Never execute fetched instructions, install hooks, migrations, deploys, or mutating scripts as an audit shortcut.
• Keep raw output in the report evidence or a referenced artifact. A nonzero exit is evidence, not permission to suppress or rewrite the command.

Manual inspection requirements

• Review accessible names in context and screen-reader reading order
• Test high-risk custom widgets against the APG pattern

Evidence requirements

• Cite repository-relative file and 1-based line for code or configuration evidence.
• Record exact command and exit code for an automated check.
• Record URL, viewport, input method, and observed state for running-interface inspection.
• Name the test and demonstrate that it exercises the claimed behavior.
• Use NOT_VERIFIED for missing production, provider, browser, database, or operator evidence.
• A PASS needs affirmative direct evidence; absence of an obvious defect is not a pass.

Finding identifiers and severity

Use IDs FF-ACCE-001, FF-ACCE-002, and so on. Preserve an ID across verification and report formats.

• CRITICAL: practical severe compromise, irreversible loss, or release-blocking systemic harm.
• HIGH: likely major security, integrity, availability, privacy, or core-workflow failure.
• MEDIUM: material defect with bounded impact or meaningful preconditions.
• LOW: localized robustness, maintainability, or user-impact defect.
• INFO: verified context or improvement with no current defect.

Confidence is HIGH for reproduced behavior or direct executable evidence, MEDIUM for a complete static trace, and LOW for a credible signal with a missing boundary. Severity and confidence are independent.

Safe automatic fixes

• Associate labels, restore visible focus, and add missing semantic state
• Add text alternatives supported by surrounding context

Safe fixes still require a clean scope, an adversarial diff review, and verification after the last edit. Never broaden --safe into an architectural or policy decision.

Risky changes requiring approval

• Replacing a custom interaction or materially changing content meaning

Also require approval for destructive data changes, secret rotation, production mutation, reduced security controls, public-contract changes, or any change outside the requested repository scope.

Verification procedure

• Re-run automated checks and keyboard walkthroughs
• Map findings to exact WCAG success criteria

Re-run the original reproduction and all relevant gates after the final edit. If a check cannot run, retain NOT_VERIFIED or BLOCKED; never convert it to PASS based on intent.

Report fields

Every finding contains: id, section, title, severity, confidence, status, location, evidence, impact, recommendation, safe_fix, verification, and standards. Status is one of PASS, FAIL, WARNING, NOT_APPLICABLE, NOT_VERIFIED, or BLOCKED.

Primary standards

• WCAG 2.2 Level AA
• WAI-ARIA 1.2
• WAI-ARIA Authoring Practices Guide

Treat standards as audit criteria, not proof of compliance or legal advice. Record the version or retrieval date for time-sensitive guidance.

Stack-specific guidance

• Prefer native HTML or platform controls before ARIA-based reimplementation

Adapt filenames and commands to detected evidence. Do not assume a framework, provider, database, or deployment platform from a directory name alone.

Known limitations

• Automated scanners cover only part of WCAG; unperformed assistive-technology tests are NOT_VERIFIED

Completion contract

Never declare a feature complete merely because code was written. A task is complete only when:

1. The requested behavior is implemented.
2. Relevant workflows work end to end.
3. Authentication and authorization are verified.
4. Database behavior is reviewed.
5. Loading, empty, error, and success states exist.
6. Applicable accessibility requirements are addressed.
7. Automated checks pass.
8. Security-sensitive changes receive security review.
9. Performance-sensitive changes receive performance review.
10. Remaining risks, skipped checks, and assumptions are reported.

Never hide failed checks or claim that an operation ran when it did not.