agentsclimarketplace

Accessibility conformance program review

Skill SylphxAI/skills/skills/accessibility-conformance-program-review

Public agent skills from SylphxAI — standards, product procedures, and one-command sync for Codex, Claude Code, and Grok Build

Install
npx -y skills add SylphxAI/skills --skill accessibility-conformance-program-review

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

2 things to look at

  • 18 days oldThe repository was created 18 days ago. New is not bad, but a brand new repository carrying a familiar-sounding name is the shape a typosquat arrives in, and there has been no time for anyone else to find a problem with it.
  • 1 stars1 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.

What its author says it does

Copied from the file, not written here

Design an accessibility conformance program: standards, AT testing, issue governance, release evidence, VPAT/ACR, remediation, and exceptions. Not one UI critique, one code fix, or legal advice.

SKILL.md

4.2 KB, as published. Nobody here has run it

Accessibility Conformance Program Review

Build an evidence-backed program that can support product decisions without turning a scanner result into a conformance claim.

Workflow

  1. Define the decision: internal readiness, release gate, procurement response, VPAT/ACR input, remediation portfolio, or claim review.
  2. Inventory in-scope products, platforms, user roles, languages, content, third-party surfaces, and representative end-to-end tasks. Record exclusions.
  3. Read references/accessibility-conformance-program-patterns.md.
  4. Run the current-authority-at-use protocol before citing standards, templates, laws, or platform requirements. Mark unresolved authority as a blocker rather than filling it from memory.
  5. Build the requirement-to-evidence matrix. Combine automated checks with keyboard, screen reader, zoom/reflow, contrast, target size, motion, captions/audio, cognitive, and platform-native testing as applicable.
  6. Triage findings by user-task impact, reach, workaround quality, recurrence, and evidence confidence. Assign remediation, prevention, and retest owners.
  7. Decide release and claim status independently. A product may ship with a recorded exception without earning an unqualified conformance claim.
  8. Produce the conformance program, claim ledger, issue portfolio, release decision, exception register, and evidence-refresh cadence.

When not to use

  • For a component or flow design critique, use the relevant interface or accessibility implementation workflow; do not create a company-wide program.
  • For source-code remediation, hand off verified findings and acceptance tests to the owning implementation agent.
  • For legal applicability or contractual interpretation, identify the question and evidence gap, then defer the conclusion to authorized counsel.
  • For generic test planning with no accessibility claim or governance decision, use the product's QA process.

Source verification

  • Resolve the exact product version, standard or platform requirement, version, jurisdiction or contract scope, retrieval time, evaluator method, assistive technology/browser/platform combination, and evidence owner at task time.
  • Prefer current primary standards, official templates, platform requirements, and signed procurement terms. Treat summaries and model memory as discovery leads only; block the affected claim when primary authority is unavailable.
  • Keep supplied test results immutable and source-linked. Never infer an untested workflow, platform, disability need, or conformance status from a neighbouring result.

Guardrails

  • Never claim conformance from automated scans alone.
  • Never claim legal compliance, certification, or a completed VPAT/ACR without scoped, current, reviewable evidence and known exceptions.
  • Never average away a blocker that prevents a disabled user from completing a core task.
  • Never treat a design-system pass as proof that composed product workflows pass.
  • Keep disability-related research data minimal, consented, and access-controlled.
  • Distinguish observed, inferred, not_tested, and blocked evidence states.

Output

Decision and scope:
- audience / release or claim decision / surfaces / workflows / exclusions

Authority ledger:
| Requirement source | Version/date | Retrieved at | Scope | Status | Evidence owner |
| --- | --- | --- | --- | --- | --- |

Requirement-to-evidence matrix:
| Requirement | Workflow/surface | Method + AT/browser/platform | Result | Confidence | Evidence |
| --- | --- | --- | --- | --- | --- |

Issue and remediation portfolio:
| Finding | User-task impact | Reach | Workaround | Owner | Due | Retest | Release effect |
| --- | --- | --- | --- | --- | --- | --- | --- |

Claim, release, and exception decisions:
- claim status / release status / caveats / exception owner / expiry

Prevention and refresh plan:
- design-system contracts / regression tests / training / evidence review cadence

Keep looking

Skills are one crate of 328,083. Ordering is by how many stacks a row turns up in, so the top of any crate is what has actually been picked rather than what has the most stars.