@wholisphere.ai/cli

wholisphere — official command-line interface for the Wholisphere API.

Install

npm install -g @wholisphere.ai/cli
# or run on demand
npx @wholisphere.ai/cli health

Configure

export WHOLISPHERE_API_KEY="whk_..."   # from the dashboard Keys tab
# Optional:
export WHOLISPHERE_BASE_URL="https://api.wholisphere.ai"   # override (e.g., staging)

Commands

# Liveness probe
wholisphere health

# Image → narration
wholisphere describe --input photo.png --url https://example.com/product/12

# Simplify text
echo 'A long technical paragraph...' | wholisphere simplify --level grade-4
wholisphere simplify --input long.txt --level plain

# Summarize a long doc
wholisphere summarize --input page.txt --max-bullets 5

# Generate a VPAT (markdown to stdout)
wholisphere vpat --input vpat-input.json > vpat.md

# Submit findings from your scan output
wholisphere findings submit \
  --product "Acme App" --version 4.2.0 \
  --build-ref "$GITHUB_SHA" \
  --input scan.json

# Invalidate the cache after a deploy
wholisphere invalidate --urls "https://example.com/*" --reason "release v4.2.0"

Output convention

• stdout — primary output (markdown, narrative text, JSON). Pipe-friendly.
• stderr — status messages, errors, latency info.

Exit codes

| Code | Meaning | |---:|---| | 0 | Success | | 1 | Unknown error | | 2 | Bad invocation (missing args, unknown command) | | 3 | Missing credentials | | 4 | API client error (4xx) | | 5 | API server / retryable error (5xx) |

VPAT input format

{
  "product": {
    "productName": "Acme App",
    "productVersion": "4.2.0",
    "vendorName": "Acme Inc.",
    "contactInfo": "[email protected]",
    "reportDate": "2026-04-18",
    "productDescription": "Online ordering for Acme products.",
    "standards": ["WCAG 2.2 Level AA", "Section 508"]
  },
  "findings": [
    { "criterion": "1.1.1", "conformance": "Supports", "remarks": "All images have alt text." }
  ]
}

Help

wholisphere --help