Packages

Everything published for The Neural Ledger, across npm, PyPI, and container registries.

npm

Package	Purpose
@theneuralledger/sdk	Typed TypeScript client for the intelligence API
@theneuralledger/cli	tnl command line tool and foreground sync daemon
@theneuralledger/mcp	Read-only MCP server (tnl-mcp) for AI clients
@theneuralledger/events	Webhook signing, verification, and receiver toolkit
@theneuralledger/research	Evidence-first research skills and orchestration
@theneuralledger/adapters	Host-neutral AI client adapter contracts
@theneuralledger/connectors	Actions, signed triggers, and lifecycle contracts for automation platforms
n8n-nodes-tnl-intelligence	n8n community node

npm install @theneuralledger/sdk

All npm packages are MIT licensed and published with provenance, so the registry records which repository and workflow built each tarball.

PyPI

Package	Purpose
tnl-intelligence	Sync and async API client, plus webhook verification

pip install tnl-intelligence

Containers

Image	Purpose
ghcr.io/bekirdag/tnl-intelligence	CLI and MCP server
ghcr.io/bekirdag/tnl-intelligence-gateway	Gateway service

docker run --rm -e TNL_API_KEY=$TNL_API_KEY \
  ghcr.io/bekirdag/tnl-intelligence tnl latest --limit 5

Which one do I want?

Reading intelligence from a script or service — the SDK for TypeScript, or tnl-intelligence for Python. Both handle auth, retries, pagination, and rate limit headers.

Exploring from a terminal — the CLI. No code required.

Connecting an AI assistant — the MCP server. See MCP overview.

Receiving pushed events — @theneuralledger/events for Node, or tnl-intelligence for Python. See verifying webhooks.

Building on an automation platform — you probably want an existing integration rather than a package.

Versioning

Packages follow semantic versioning and are released together, so versions stay aligned across the set. Pin exact versions in production and upgrade deliberately — a minor release can add fields to responses.