a11y-core
Shared contract for the Accessibility Agents skills - dispatch, findings schema, report rules. Read by skills, never dispatched…
Open Source Accessibility Framework Foundational phase: ACCESSIBILITY.md and an accessibility label.
Claude Code
$ npx -y skills add Community-Access/accessibility-agents --skill osaf-foundational --agent claude-code
Codex
$ npx skills add Community-Access/accessibility-agents --skill osaf-foundational -a codex -y
Cursor
$ npx skills add Community-Access/accessibility-agents --skill osaf-foundational -a cursor -y
Gemini CLI
$ npx skills add Community-Access/accessibility-agents --skill osaf-foundational -a gemini-cli -y
GitHub Copilot
$ npx skills add Community-Access/accessibility-agents --skill osaf-foundational -a github-copilot -y
OpenCode
$ npx skills add Community-Access/accessibility-agents --skill osaf-foundational -a opencode -y
Windsurf
$ npx skills add Community-Access/accessibility-agents --skill osaf-foundational -a windsurf -y
name: osaf-foundational description: "Open Source Accessibility Framework Foundational phase: ACCESSIBILITY.md and an accessibility label." license: MIT disable-model-invocation: true metadata: tier: specialist domain: github output: artifact effort: medium title: "OSAF Foundational Phase" osaf-phase: foundational osaf-commit: "b2ea31a2e38cbab3e774e4142575b78876b9b53f" osaf-digest: "d6f531edfd4e" osaf-source: "https://raw.githubusercontent.com/open-source-accessibility/Open-Source-Accessibility-Framework/main/framework/phases/foundational-phase.md"
You implement the Foundational phase of the Open Source Accessibility Framework in one repository: both of its actions, or the one the request names. You write what the actions need, verify each against its Definition of Done, and return the work for the maintainer to review. This phase comes before the other three. Framework facts, the full action index and the tracking issue: `kb-open-source-accessibility-framework`.
1. **Scope.** Every action in phase order, or only those named by number, title or ID. If the request names none, ask before changing files. If a named action does not exist, say so and list the valid ones below. 2. **Bundled requirements.** Each Definition of Done below is the minimum, taken from the canonical phase document at `metadata.osaf-commit`. It works offline. Read an action's reference only when that action is in scope. 3. **Online check, only when asked** for the latest guidance. Fetch `metadata.osaf-source`, take the first 12 hex characters of its SHA-256 and compare with `metadata.osaf-digest`. If they differ, report what changed and ask before using it. A failed fetch never blocks the work. 4. **Inspect before editing.** For each Definition of Done item decide: met (cite the file and line), needs a change, or needs maintainer input, external configuration or a person. 5. **Edit in phase order.** Keep the repository's conventions. Extend an existing accessibility document rather than replace it. 6. **Ask before public changes.** Creating or applying labels and opening the tracking issue change GitHub, not files. Prepare the commands; run them only on an explicit yes and with the access they need. 7. **Verify** each action: met, unmet, recurring work (action 2 is ongoing), and follow-up for a person or an external change. 8. **Present for review** before anything is committed, pushed or published. 9. **No compliance claims.** Never say completing the phase makes the project accessible, compliant, conformant or certified. The framework is for improvement, not certification.
When the request is to adopt the framework as a whole, also prepare the tracking issue: give the user the form link and the field values the reference skill lists. Opening it publishes the project's participation, so it is the user's call.
Each action reference names the skills in this package that do part of the work better. Return those as hand-offs rather than dispatching them; `github-hub` dispatches the GitHub ones and passes markdown review to `markdown-a11y-assistant`.
`create-an-accessibilitymd` - Quick - One-time setup - `references/action-01.md`
1. An `ACCESSIBILITY.md` exists in the repository. 2. The file includes a brief statement describing why accessibility matters to your project.
`use-an-accessibility-label-to-track-relevant-work` - Quick - Initial setup + ongoing - `references/action-02.md`
1. An `accessibility` label, or an equivalent label with a clear name, exists. 2. The label is applied to the project's relevant open tracking issues.
Write the files the actions need. Then report, in this order:
1. Files written or changed, and what each change does. 2. Per action, each Definition of Done item as met, unmet or needs a person, with evidence: file and line, command output, or issue numbers. 3. Hand-offs: skill, task and scope, one per line. 4. Tracking issue: the checklist lines that can now be checked. Do not edit the tracking issue yourself.
The shape is `examples/sample-output.md`. Adapted framework text is MIT; the notice is `references/LICENSE-OSAF.md`.
Shared rules, dispatch contract and schemas: `skills/a11y-core/SKILL.md`. Authoritative specifications for this skill: `skills/a11y-core/references/sources.md`.
WCAG 2.2 AA enforcement for agentic coding, as a set of Agent Skills. One package, read natively by Claude Code, Codex, GitHub Copilot, Gemini CLI and Antigravity, with no per-client copies. Models forget accessibility while generating code.
Shared contract for the Accessibility Agents skills - dispatch, findings schema, report rules. Read by skills, never dispatched…
Build accessibility scanners, rule engines, parsers and report generators.
Web UI accessibility lead. Use before writing or changing HTML, JSX, TSX, Vue, Svelte, CSS or templates. Picks specialists and…
Compare audits across commits to find new, fixed and regressed issues.
Generate a W3C or EU model accessibility statement from audit results.
GitHub Actions: workflow runs, logs, re-runs and CI failure triage.
Produce the "Playwright 1.xx API Review" artifact for a range of commits - the public API additions and the breaking or behavior…
Write, refine, run, and QA promptfoo evaluation suites: promptfooconfig.yaml, prompts, providers, vars, tests, assertions,…
DeepEval evaluation workflow for AI agents and LLM applications. TRIGGER when the user wants to evaluate or improve an AI agent,…