td-ai-tools 1.1.7 → 1.1.9

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "td-ai-tools",
-  "version": "1.1.7",
+  "version": "1.1.9",
   "description": "Install agent skills and packs into your project",
   "type": "module",
   "scripts": {

package/skills/README.md

@@ -1,7 +1,10 @@
 # skills
 
 ## Available Skills
+- `a11y-audit`: Run script-driven accessibility audits for user-provided URLs with Axe Core and Playwright, targeting…
+- `barrage`: Convert a group of Basecamp todos or cards into a td-barrage queue.json task file.
 - `basecamp`: Interact with Basecamp via the Basecamp CLI.
+- `browser-validation`: Before completing a task validate frontend or template changes in a real browser with the Playwright-CLI…
 - `cache-reset`: Clear and warm Laravel and Statamic caches (including Statamic Glide image caches) after content or template…
 - `car-ticket-generator`: Generate a ticket for the codex-auto-runner queue
 - `everhour-basecamp-estimates`: Bulk update Everhour task estimates from a Basecamp todo or todolist URL, then append bracketed hours to the…
@@ -12,6 +15,7 @@
 - `pull-request`: Generates well-structured GitHub pull request descriptions for Shopify theme development teams by analyzing…
 - `pull-request-statamic`: Generates GitHub pull request descriptions for Statamic and Laravel development by analyzing git diffs and…
 - `record-changes`: Update `docs/changes.md` by summarizing the current branch against the primary development branch.
+- `scry`: Single-site visual regression workflow for comparing a live URL against a preview/staging URL in Playwright…
 - `stylesheet-migration`: Migrate Shopify Liquid `{% stylesheet %}` blocks into theme CSS assets using bundled Python scripts.
 - `td-js-vanilla-rules`: Theory Digital vanilla JavaScript standards for Shopify theme work.
 - `td-review`: Run parallel code review agents on a PR (including TD theme compliance) and produce a synthesized findings…

package/skills/a11y-audit/SKILL.md

@@ -0,0 +1,60 @@
+---
+name: a11y-audit
+description: Run script-driven accessibility audits for user-provided URLs with Axe Core and Playwright, targeting baseline or WCAG thresholds such as WCAG 2.1 AA or WCAG 2.2 AA. Use when Codex is asked to audit a page/site for accessibility, produce a prioritized list of required accessibility changes, identify WCAG/Axe violations, or perform manual accessibility review with playwright-cli.
+---
+
+# A11y Audit
+
+## Workflow
+
+1. Run the Axe audit script first. Keep dependencies local to this skill directory.
+
+```bash
+cd .agents/skills/a11y-audit
+npm run setup
+node scripts/audit.js "https://example.com" --threshold wcag22aa
+```
+
+If the user does not name a threshold, use `baseline`. Accepted thresholds: `baseline`, `wcag2a`, `wcag2aa`, `wcag21aa`, `wcag22aa`. The script also accepts multiple URLs and `--tags` for custom Axe tag filters.
+
+2. Read the generated Markdown report path printed by the script. Use it as the primary source of automated findings and required changes.
+
+3. If the report includes `Manual review required`, review those items with `$playwright-cli`. Use browser snapshots, keyboard navigation, screenshots, and targeted page evaluation as needed. Do not guess outcomes for manual-only criteria.
+
+4. Return the final audit as a concise remediation list:
+- Automated failures from Axe that must be fixed.
+- Manual-review failures confirmed with `playwright-cli`.
+- Manual checks that could not be completed, with the blocker and exact next action.
+- Threshold tested, URL(s), viewport(s), and report artifact paths.
+
+## Script
+
+Use `scripts/audit.js` for deterministic collection. It:
+- launches Chromium through Playwright;
+- injects Axe Core from the skill-local dependency;
+- audits desktop and mobile viewports by default;
+- writes `report.md`, `results.json`, and `manual-review.json`;
+- groups issues by impact, rule, WCAG tags, affected nodes, and suggested remediation;
+- lists Axe `incomplete` results and threshold-specific human checks for manual review.
+
+Exit code `2` means the audit completed and found violations or manual-review items. Treat it as a findings signal, not a runner failure.
+
+Useful options:
+
+```bash
+node scripts/audit.js URL [URL ...]
+node scripts/audit.js URL --threshold wcag22aa --out ./a11y-report
+node scripts/audit.js URL --viewport desktop --viewport 390x844
+node scripts/audit.js URL --tags wcag2a,wcag2aa,wcag21aa,wcag22aa
+node scripts/audit.js URL --wait-until networkidle --timeout 45000
+```
+
+## Manual Review With playwright-cli
+
+Use `playwright-cli` only after the script flags manual review or the user asks for deeper investigation. Typical checks:
+- Keyboard path: `Tab`, `Shift+Tab`, `Enter`, `Space`, `Escape`; confirm logical order, visible focus, no traps, and operable controls.
+- Screen-reader semantics proxy: inspect snapshots for names, roles, headings, landmarks, labels, and state changes.
+- Visual checks: screenshots for focus indication, responsive reflow, text spacing, content on hover/focus, contrast edge cases, and motion/animation behavior.
+- Forms and dynamic UI: submit invalid states, open menus/dialogs, expand accordions, trigger errors, and verify announcements or visible status messages.
+
+Close browser sessions when finished.

package/skills/a11y-audit/agents/openai.yaml

@@ -0,0 +1,4 @@
+interface:
+  display_name: "A11y Audit"
+  short_description: "Run Axe Core audits with manual review guidance."
+  default_prompt: "Use $a11y-audit to audit the provided URL against WCAG 2.2 AA."

package/skills/a11y-audit/package.json

@@ -0,0 +1,17 @@
+{
+  "name": "a11y-audit-skill",
+  "version": "1.0.0",
+  "private": true,
+  "description": "Local dependencies for the a11y-audit Codex skill.",
+  "license": "UNLICENSED",
+  "type": "commonjs",
+  "scripts": {
+    "audit": "node scripts/audit.js",
+    "setup": "bash scripts/setup.sh"
+  },
+  "dependencies": {
+    "axe-core": "^4.10.3",
+    "playwright": "^1.52.0"
+  },
+  "devDependencies": {}
+}