Content Intelligence

POST https://kovca.io/api/v1/content/moderate

Authenticate with your API key in the Authorization: Bearer header. Send JSON with Content-Type: application/json; bodies are limited to 16 KB.

The API evaluates one piece of content against a small set of default categories and the rules you send. It blocks only when your own rules or block_categories support it; default categories alone never block. A category describes the text, never the author's intent or real-world behavior. Content is not stored, and the API never publishes, deletes, or suspends anything.

Request body

{
  "content": "Great blender but you can get it 40% cheaper at bestdeals-example.shop, just message me on WhatsApp +1 555 0100.",
  "content_type": "product_review",
  "context": "Marketplace product reviews",
  "policy": {
    "disallowed": [
      "Links or contact details that take buyers off the platform",
      "Reviews of a different product than the one listed"
    ],
    "allowed": [
      "Negative reviews and criticism of products or service"
    ],
    "block_categories": [
      "spam",
      "threats",
      "harassment"
    ]
  }
}

Response

{
  "request_id": "req_Lt6qN3wRv8KzM1pY",
  "decision": "block",
  "recommended_action": "block_submission",
  "categories": [
    {
      "category": "spam",
      "assessment": "likely"
    },
    {
      "category": "policy_violation",
      "assessment": "likely"
    }
  ],
  "policy_matches": [
    {
      "rule": "Links or contact details that take buyers off the platform",
      "source": "policy.disallowed[0]",
      "assessment": "likely"
    }
  ],
  "signals": [
    "custom_rule_matched"
  ]
}

Decisions (first match wins)

• insufficient_information: Fewer than 3 letters or digits. Action: request_information.
• block: A likely match on a concrete disallowed rule, or a likely category listed in block_categories. Action: block_submission.
• review: Any other match, including likely default categories you didn't list in block_categories, and any possible match. Action: hold_for_review.
• allow: Nothing matched. Action: publish.

Default categories

• spam: Unsolicited bulk or repetitive promotion, link bait, scams, or text unrelated to the context posted to drive traffic.
• harassment: Personal attacks, slurs, or degrading language aimed at someone.
• threats: A statement of intent to hurt, attack, or damage, including veiled threats.
• sexual_content: Explicit sexual descriptions or solicitation.
• violent_content: Graphic descriptions of injury or death, or praise of violent acts.
• prohibited_promotion: It offers, advertises, or solicits illegal goods or services.

Criticism without a policy

{
  "content": "This company's support team is useless. I waited three weeks and nobody fixed my account. Avoid.",
  "content_type": "product_review"
}

Code

curl https://kovca.io/api/v1/content/moderate \
  -H "Authorization: Bearer $KOVCA_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
  "content": "Great blender but you can get it 40% cheaper at bestdeals-example.shop, just message me on WhatsApp +1 555 0100.",
  "content_type": "product_review",
  "context": "Marketplace product reviews",
  "policy": {
    "disallowed": [
      "Links or contact details that take buyers off the platform",
      "Reviews of a different product than the one listed"
    ],
    "allowed": [
      "Negative reviews and criticism of products or service"
    ],
    "block_categories": [
      "spam",
      "threats",
      "harassment"
    ]
  }
}'

Errors

• 400 INVALID_REQUEST: a field is missing, too long, the wrong type, or not allowed. The message names the field.
• 413 PAYLOAD_TOO_LARGE: the body is larger than 16 KB.
• 401 UNAUTHORIZED, 429 RATE_LIMITED, 503 SERVICE_UNAVAILABLE, 504 TIMEOUT: see Errors and limits.

Limitations

• Default criteria are limited to the six categories below. Anything else needs your rules.
• Subjective rules such as 'off-topic' or 'inappropriate' can send content to review but never block it on their own.
• Text only; no images, audio, or links are followed.
• Advisory: it recommends an action and never enforces it.
• Request data is processed to produce the result and is not stored by Kovca.