Prompt file imported from Community-Access/accessibility-agents (.github/prompts/markdown-scanner.prompt.md). Fill in {{mdFile}} before use. Copyright stays with the author.

Markdown Accessibility Scanner

Name: scan-markdown
Author: Community-Access

Scan a single markdown file for accessibility issues across all 9 domains.

File to Scan

File: {{mdFile}}

Instructions

Step 1: Read Markdown File

Parse file content, track line numbers.

Step 2: Scan Domain 1 - Links

Ambiguous link text patterns:

"click here", "here", "link"
"read more", "learn more", "more info"
"this", "this page", "this site"

Check:

Link text descriptive?
Context provides clarity?

WCAG: 2.4.4 Link Purpose (In Context)

Step 3: Scan Domain 2 - Alt Text

Find all image references:

![alt text](image.png)

Check:

Alt text present?
Alt text descriptive (not filename)?
Decorative images use empty alt ![](...)?

WCAG: 1.1.1 Non-text Content

Step 4: Scan Domain 3 - Headings

Check:

Heading hierarchy (H1 → H2 → H3, no skipping)
Only one H1 per document?
Headings descriptive?

WCAG: 1.3.1 Info and Relationships, 2.4.6 Headings and Labels

Step 5: Scan Domain 4 - Tables

Find all tables:

| Header | Header |
|--------|--------|
| Data   | Data   |

Check:

Table description present? 
Complex tables have caption?

WCAG: 1.3.1 Info and Relationships

Step 6: Scan Domain 5 - Emoji

Check:

Emoji in headings? (screen reader reads Unicode name)
Emoji as content substitute? (no text equivalent)

Options:

Remove emoji
Translate to English 🎉 → (celebrate)

WCAG: 1.1.1 Non-text Content (text alternative)

Step 7: Scan Domain 6 - Diagrams

Find Mermaid diagrams:

```mermaid
graph TD
  A --> B
```text

Find ASCII diagrams (boxes, arrows)

Check:

Text alternative present?
Diagram described in prose?

WCAG: 1.1.1 Non-text Content

Suggested fix: Provide full text description before/after diagram

Step 8: Scan Domain 7 - Anchor Links

Find anchor references:

[Jump to section](#section-name)

Check:

Target heading exists?
Anchor ID matches heading text (lowercase, dashes)?

WCAG: 2.4.1 Bypass Blocks (for navigation)

Step 9: Scan Domain 8 - Em-Dashes

Find em-dashes: — (U+2014)

Issue: Some markdown parsers convert — incorrectly

Fix: Use triple-dash ---

WCAG: 1.3.1 Info and Relationships (content stability)

Step 10: Scan Domain 9 - Quality

Check:

Code blocks have language identifiers?
Lists use consistent markers?
No trailing whitespace?

Step 11: Return Structured Findings

{
  "file": "{{mdFile}}",
  "findings": [
    {
      "domain": "links",
      "ruleId": "MD-LINK-001",
      "severity": "serious",
      "line": 42,
      "description": "Ambiguous link text: 'click here'",
      "currentValue": "[click here](https://example.com)",
      "suggestedFix": "[view the API documentation](https://example.com)",
      "autoFix": false,
      "wcag": "2.4.4"
    },
    {
      "domain": "emoji",
      "ruleId": "MD-EMOJI-001",
      "severity": "moderate",
      "line": 12,
      "description": "Emoji in heading",
      "currentValue": "# Heading with emoji 🎉",
      "suggestedFix": "# Heading without emoji (celebrate)",
      "autoFix": true,
      "wcag": "1.1.1"
    }
  ],
  "summary": {
    "critical": 0,
    "serious": 3,
    "moderate": 5,
    "minor": 2
  }
}

Expected Output

Structured JSON findings
Per-domain scan results
Auto-fix classification
Ready for markdown-fixer consumption

# Markdown Accessibility Scanner Scan a single markdown file for accessibility issues across all 9 domains. ## File to Scan **File:** `{{mdFile}}` ## Instructions ### Step 1: Read Markdown File Parse file content, track line numbers. ### Step 2: Scan Domain 1 - Links **Ambiguous link text patterns:** - "click here", "here", "link" - "read more", "learn more", "more info" - "this", "this page", "this site" **Check:** - Link text descriptive? - Context provides clarity? **WCAG:** 2.4.4 Link Purpose (In Context) ### Step 3: Scan Domain 2 - Alt Text Find all image references: ```markdown ![alt text](image.png) ``` **Check:** - Alt text present? - Alt text descriptive (not filename)? - Decorative images use empty alt `![](...)`? **WCAG:** 1.1.1 Non-text Content ### Step 4: Scan Domain 3 - Headings **Check:** - Heading hierarchy (H1 → H2 → H3, no skipping) - Only one H1 per document? - Headings descriptive? **WCAG:** 1.3.1 Info and Relationships, 2.4.6 Headings and Labels ### Step 5: Scan Domain 4 - Tables Find all tables: ```markdown | Header | Header | |--------|--------| | Data | Data | ``` **Check:** - Table description present? `` - Complex tables have caption? **WCAG:** 1.3.1 Info and Relationships ### Step 6: Scan Domain 5 - Emoji **Check:** - Emoji in headings? (screen reader reads Unicode name) - Emoji as content substitute? (no text equivalent) **Options:** - Remove emoji - Translate to English `🎉` → `(celebrate)` **WCAG:** 1.1.1 Non-text Content (text alternative) ### Step 7: Scan Domain 6 - Diagrams Find Mermaid diagrams: ````markdown ```mermaid graph TD A --> B ```text ```` Find ASCII diagrams (boxes, arrows) **Check:** - Text alternative present? - Diagram described in prose? **WCAG:** 1.1.1 Non-text Content **Suggested fix:** Provide full text description before/after diagram ### Step 8: Scan Domain 7 - Anchor Links Find anchor references: ```markdown [Jump to section](#section-name) ``` **Check:** - Target heading exists? - Anchor ID matches heading text (lowercase, dashes)? **WCAG:** 2.4.1 Bypass Blocks (for navigation) ### Step 9: Scan Domain 8 - Em-Dashes Find em-dashes: `—` (U+2014) **Issue:** Some markdown parsers convert `—` incorrectly **Fix:** Use triple-dash `---` **WCAG:** 1.3.1 Info and Relationships (content stability) ### Step 10: Scan Domain 9 - Quality **Check:** - Code blocks have language identifiers? - Lists use consistent markers? - No trailing whitespace? ### Step 11: Return Structured Findings ```json { "file": "{{mdFile}}", "findings": [ { "domain": "links", "ruleId": "MD-LINK-001", "severity": "serious", "line": 42, "description": "Ambiguous link text: 'click here'", "currentValue": "[click here](https://example.com)", "suggestedFix": "[view the API documentation](https://example.com)", "autoFix": false, "wcag": "2.4.4" }, { "domain": "emoji", "ruleId": "MD-EMOJI-001", "severity": "moderate", "line": 12, "description": "Emoji in heading", "currentValue": "# Heading with emoji 🎉", "suggestedFix": "# Heading without emoji (celebrate)", "autoFix": true, "wcag": "1.1.1" } ], "summary": { "critical": 0, "serious": 3, "moderate": 5, "minor": 2 } } ``` ## Expected Output - Structured JSON findings - Per-domain scan results - Auto-fix classification - Ready for markdown-fixer consumption

{ "prompt": "# Markdown Accessibility Scanner\n\nScan a single markdown file for accessibility issues across all 9 domains.\n\n## File to Scan\n\n**File:** `{{mdFile}}`\n\n## Instructions\n\n### Step 1: Read Markdown File\n\nParse file content, track line numbers.\n\n### Step 2: Scan Domain 1 - Links\n\n**Ambiguous link text patterns:**\n\n- \"click here\", \"here\", \"link\"\n- \"read more\", \"learn more\", \"more info\"\n- \"this\", \"this page\", \"this site\"\n\n**Check:**\n\n- Link text descriptive?\n- Context provides clarity?\n\n**WCAG:** 2.4.4 Link Purpose (In Context)\n\n### Step 3: Scan Domain 2 - Alt Text\n\nFind all image references:\n\n```markdown\n![alt text](image.png)\n```\n\n**Check:**\n\n- Alt text present?\n- Alt text descriptive (not filename)?\n- Decorative images use empty alt `![](...)`?\n\n**WCAG:** 1.1.1 Non-text Content\n\n### Step 4: Scan Domain 3 - Headings\n\n**Check:**\n\n- Heading hierarchy (H1 → H2 → H3, no skipping)\n- Only one H1 per document?\n- Headings descriptive?\n\n**WCAG:** 1.3.1 Info and Relationships, 2.4.6 Headings and Labels\n\n### Step 5: Scan Domain 4 - Tables\n\nFind all tables:\n\n```markdown\n| Header | Header |\n|--------|--------|\n| Data | Data |\n```\n\n**Check:**\n\n- Table description present? ``\n- Complex tables have caption?\n\n**WCAG:** 1.3.1 Info and Relationships\n\n### Step 6: Scan Domain 5 - Emoji\n\n**Check:**\n\n- Emoji in headings? (screen reader reads Unicode name)\n- Emoji as content substitute? (no text equivalent)\n\n**Options:**\n\n- Remove emoji\n- Translate to English `🎉` → `(celebrate)`\n\n**WCAG:** 1.1.1 Non-text Content (text alternative)\n\n### Step 7: Scan Domain 6 - Diagrams\n\nFind Mermaid diagrams:\n\n````markdown\n```mermaid\ngraph TD\n A --> B\n```text\n````\n\nFind ASCII diagrams (boxes, arrows)\n\n**Check:**\n\n- Text alternative present?\n- Diagram described in prose?\n\n**WCAG:** 1.1.1 Non-text Content\n\n**Suggested fix:** Provide full text description before/after diagram\n\n### Step 8: Scan Domain 7 - Anchor Links\n\nFind anchor references:\n\n```markdown\n[Jump to section](#section-name)\n```\n\n**Check:**\n\n- Target heading exists?\n- Anchor ID matches heading text (lowercase, dashes)?\n\n**WCAG:** 2.4.1 Bypass Blocks (for navigation)\n\n### Step 9: Scan Domain 8 - Em-Dashes\n\nFind em-dashes: `—` (U+2014)\n\n**Issue:** Some markdown parsers convert `—` incorrectly\n\n**Fix:** Use triple-dash `---`\n\n**WCAG:** 1.3.1 Info and Relationships (content stability)\n\n### Step 10: Scan Domain 9 - Quality\n\n**Check:**\n\n- Code blocks have language identifiers?\n- Lists use consistent markers?\n- No trailing whitespace?\n\n### Step 11: Return Structured Findings\n\n```json\n{\n \"file\": \"{{mdFile}}\",\n \"findings\": [\n {\n \"domain\": \"links\",\n \"ruleId\": \"MD-LINK-001\",\n \"severity\": \"serious\",\n \"line\": 42,\n \"description\": \"Ambiguous link text: 'click here'\",\n \"currentValue\": \"[click here](https://example.com)\",\n \"suggestedFix\": \"[view the API documentation](https://example.com)\",\n \"autoFix\": false,\n \"wcag\": \"2.4.4\"\n },\n {\n \"domain\": \"emoji\",\n \"ruleId\": \"MD-EMOJI-001\",\n \"severity\": \"moderate\",\n \"line\": 12,\n \"description\": \"Emoji in heading\",\n \"currentValue\": \"# Heading with emoji 🎉\",\n \"suggestedFix\": \"# Heading without emoji (celebrate)\",\n \"autoFix\": true,\n \"wcag\": \"1.1.1\"\n }\n ],\n \"summary\": {\n \"critical\": 0,\n \"serious\": 3,\n \"moderate\": 5,\n \"minor\": 2\n }\n}\n```\n\n## Expected Output\n\n- Structured JSON findings\n- Per-domain scan results\n- Auto-fix classification\n- Ready for markdown-fixer consumption", "variables": [ "mdFile" ], "notes": "Single markdown file accessibility scanner. Scans across all 9 domains (links, alt, headings, tables, emoji, diagrams, anchors, em-dashes, quality). Returns structured findings. Mode: agent. Tools: askQuestions, readFile.", "metadata": { "source": { "provider": "github-prompts", "repository": "https://github.com/Community-Access/accessibility-agents", "path": ".github/prompts/markdown-scanner.prompt.md", "ref": "161c60c7493ad657f371ad8f91253d33c3b12044", "url": "https://github.com/Community-Access/accessibility-agents/blob/161c60c7493ad657f371ad8f91253d33c3b12044/.github/prompts/markdown-scanner.prompt.md", "key": "Community-Access/accessibility-agents/.github/prompts/markdown-scanner.prompt.md" } } }

scan-markdown

About

Markdown Accessibility Scanner

File to Scan

Instructions

Step 1: Read Markdown File

Step 2: Scan Domain 1 - Links

Step 3: Scan Domain 2 - Alt Text

Step 4: Scan Domain 3 - Headings

Step 5: Scan Domain 4 - Tables

Step 6: Scan Domain 5 - Emoji

Step 7: Scan Domain 6 - Diagrams

Step 8: Scan Domain 7 - Anchor Links

Step 9: Scan Domain 8 - Em-Dashes

Step 10: Scan Domain 9 - Quality

Step 11: Return Structured Findings

Expected Output

Use it

Manifest

Reviews