`search_incidents`

Pack: ai-incident-db · Endpoint: https://gateway.pipeworx.io/ai-incident-db/mcp

Search the AI Incident Database. Combines free-text with date range filters. Returns incident IDs, titles, dates, and a short blurb.

Parameters

Name	Type	Required	Description
`query`	string	no	Free-text. Every word must appear somewhere in the incident title or description, in any order — “facial recognition bias” finds records mentioning all three, not that exact phrase. Fewer words match more.
`start_date`	string	no	YYYY-MM-DD lower bound on incident date
`end_date`	string	no	YYYY-MM-DD upper bound
`limit`	number	no	1-200 (default 25)
`offset`	number	no	0-based offset

Example call

Arguments

{
  "query": "facial recognition bias"
}

curl

curl -X POST https://gateway.pipeworx.io/ai-incident-db/mcp \
  -H 'Content-Type: application/json' \
  -d '{"jsonrpc":"2.0","id":1,"method":"tools/call","params":{"name":"search_incidents","arguments":{"query":"facial recognition bias"}}}'

TypeScript (@pipeworx/sdk)

import { Pipeworx } from '@pipeworx/sdk';
const pipeworx = new Pipeworx();

const result = await pipeworx.call('search_incidents', {
  "query": "facial recognition bias"
});

More examples

{
  "query": "chatbot misinformation",
  "start_date": "2024-01-01",
  "end_date": "2025-01-01",
  "limit": 50
}

Response shape

Field	Type	Description
`incidents`	array	List of matching incidents
`returned`	number	How many incidents are in this page (<= the requested limit).
`offset`	number	The offset this page started at, echoed back.
`has_more`	boolean	True if more incidents exist past this page. AIID exposes no total count, so this is the only way to distinguish “end of results” from an empty response.

Full JSON Schema

{
  "type": "object",
  "properties": {
    "incidents": {
      "type": "array",
      "items": {
        "type": "object",
        "properties": {
          "incident_id": {
            "type": "number",
            "description": "Unique incident identifier"
          },
          "title": {
            "type": "string",
            "description": "Incident title"
          },
          "date": {
            "type": "string",
            "description": "Incident date in YYYY-MM-DD format"
          },
          "description": {
            "type": "string",
            "description": "Short description of the incident"
          },
          "editors_notes": {
            "type": "string",
            "description": "Editorial notes on the incident"
          }
        }
      },
      "description": "List of matching incidents"
    },
    "returned": {
      "type": "number",
      "description": "How many incidents are in this page (<= the requested limit)."
    },
    "offset": {
      "type": "number",
      "description": "The offset this page started at, echoed back."
    },
    "has_more": {
      "type": "boolean",
      "description": "True if more incidents exist past this page. AIID exposes no total count, so this is the only way to distinguish \"end of results\" from an empty response."
    }
  }
}

Connect

Add this to your MCP client config, or use one-click install buttons:

{
  "mcpServers": {
    "ai-incident-db": {
      "url": "https://gateway.pipeworx.io/ai-incident-db/mcp"
    }
  }
}

See Getting Started for client-specific install steps.

search_incidents