Self-Healing Scraper & DOM-to-JSON Compiler

Synthesizes durable CSS selectors (Cheerio; no XPath) for a target JSON schema, proves them against the live DOM in-request, and returns the extracted data plus a ready-to-run TypeScript Cheerio extractor.

Every result is a signed, verifiable answer:

• a W3C Verifiable Credential 2.0, signed with eddsa-jcs-2022 by did:web:agents.writerslogic.com;
• a C2PA-style manifest that lists every source the agent read, each with its SHA-256;
• an EIP-191 co-signature over the same digest.

Change one byte and verification fails, so you can pass the result to a client, an auditor or another agent and they can check it without trusting you or us.

How to use Self-Healing Scraper & DOM-to-JSON Compiler

1. Fill in the input form. Required fields are marked; the form opens with a working example.
2. Start the run.
3. Read the result in the Output tab, or download the dataset as JSON. The full signed envelope is also saved as the OUTPUT record in the key-value store.

Input

Field	Required	Description
url	no	Page to fetch. Provide url or html.
html	no	Raw HTML to use instead of fetching.
targetSchema	yes	JSON Schema of the fields to extract: an object whose properties are string/number/integer/boolean (or arrays of those), or an array whose items are such an object for list pages. The pattern keyword is not allowed.
previousBrokenSelectors	no	Map of field name to the CSS selector that stopped working.

Example input:

{
  "url": "https://news.ycombinator.com",
  "targetSchema": {
    "type": "array",
    "items": {
      "type": "object",
      "required": [
        "title",
        "link"
      ],
      "properties": {
        "title": {
          "type": "string"
        },
        "link": {
          "type": "string"
        },
        "points": {
          "type": "integer"
        }
      }
    }
  },
  "previousBrokenSelectors": {
    "title": ".storylink"
  }
}

Output

{
  "agent": "heal-scraper",
  "priceUsd": 0.50,
  "data": { "...": "the agent's answer" },
  "_provenance": { "credential": {}, "manifest": {}, "ethSignature": "0x..." }
}

To check a result for free, POST { "data": ..., "_provenance": ... } to https://agents.writerslogic.com/api/v1/verify-provenance, or verify it offline against the public key in the DID document.

Pricing

$0.50 per successful run, charged once and only after the agent returns its answer. If the input is rejected or the agent fails, you are not charged for the call.

Use it from code or an AI assistant

• Apify API and integrations: call this Actor from Make, Zapier, n8n, LangChain or the Apify API like any other Actor.
• MCP: npx -y @writerslogic/ai-agents-mcp exposes this agent and the rest of the hub as tools.
• x402: the same agent is available at POST https://agents.writerslogic.com/api/v1/heal-scraper, paid per call in USDC on Base.

More verifiable agents

• Cross-Lingual Dark Science Arbitrage Scout
• Scientific Consensus & Replication Failure Auditor
• Forensic AI-trajectory & prose craft auditor
• Crosstalk epistemic swarm & falsification engine
• Open Research Opportunity Miner
• Bibliography Reference Verifier
• All agents in one Actor

FAQ

Who runs this? WritersLogic. The agent runs on the WritersLogic Verifiable AI Agent Hub; this Actor sends your input there and returns the signed result.

Is my input stored? The Actor stores your result in your own Apify storage. Ask before sending confidential documents; the hub is a third-party service from your point of view.