TNL Intelligence Actor

Run cited TNL Intelligence search, monitoring, exposure, and research workflows on Apify. The Actor uses the public TNL API and MCP service; it does not scrape The Neural Ledger website.

What it does

Choose one operation per run:

• Search intelligence with filters and cursor pagination.
• Retrieve one intelligence item with revision and source metadata.
• List published, revised, or retracted intelligence since a timestamp.
• Find entity, asset, or impact-path exposure.
• Run one of six evidence-first TNL research workflows.
• Retrieve the cited weekly consequential-development edition.

Results are written to the default Apify dataset. Story rows preserve stable IDs, revisions, publication/update times, canonical URLs, citations, entities, assets, and impact paths. Research rows preserve the complete structured result and its citations.

Authentication

Create a TNL API key in your account at The Neural Ledger. Enter it in the TNL API key field. The field is marked as an Apify secret. The Actor never writes the key to datasets, key-value stores, or logs.

Do not put a key into a public task, README, source file, or unmasked environment variable. Revoke the key immediately if it is ever exposed.

Input examples

Search intelligence

{
  "operation": "search_intelligence",
  "apiKey": "YOUR_TNL_API_KEY",
  "query": "supply-chain disruptions affecting semiconductor manufacturing",
  "pageSize": 20,
  "includeBody": false
}

Run cited research

{
  "operation": "run_research",
  "apiKey": "YOUR_TNL_API_KEY",
  "workflowId": "operational-risk",
  "question": "What changed in Red Sea shipping risk and which operations are exposed?",
  "depth": "standard"
}

Only fields used by the selected operation are sent to TNL. Keep pageSize between 1 and 100. Use ISO 8601 timestamps for all date fields.

Output

Page operations create one dataset row per intelligence item. The OUTPUT record in the default key-value store contains the operation, item count, completion time, and the next pagination cursor when one is available.

Research operations create one dataset row containing a concise summary, research status, citations, and the complete structured result in payload.

Limits and errors

Your TNL subscription controls available scopes, quotas, and research access. The Actor retries temporary MCP failures once. Invalid/revoked credentials, quota exhaustion, invalid inputs, and unavailable upstream services produce sanitized errors that do not echo credentials or upstream response bodies.

Privacy and support

The Actor sends the fields needed for the selected operation and the API key to TNL over HTTPS. Apify stores run input and output according to the storage and retention settings of the Apify account that starts the run. Review the TNL Privacy Policy and TNL Terms.

For support, use the TNL Intelligence GitHub issue tracker or contact tnladmin@theneuralledger.com. Security reports should follow the security policy.

Source and license

Source: bekirdag/tnl-intelligence

License: MIT