AI Agent Docs

Use Lodger documentation, SDKs, and HTTP references from AI coding agents safely.

Lodger docs are structured so coding agents can implement integrations without guessing endpoint shapes or event semantics.

Agent-ready files

When hosted on Mintlify, the docs expose:

• /llms.txt for the documentation index.
• /llms-full.txt for full-site context.
• /skill.md for Lodger capabilities and integration constraints.
• /.well-known/agent-skills/index.json for agent skill discovery.

Agents should start with llms.txt, then read the specific guide or endpoint page needed for the task.

Safe integration boundaries

Agents should follow these rules when adding Lodger to an app:

• Use the TypeScript SDK from server-side code.
• Never put LODGER_API_KEY in public browser bundles.
• Prefer a stable event catalog over ad hoc event names.
• Add idempotency keys for any retryable write.
• Include request, session, trace, environment, and server context when available.
• Store returned receipt IDs next to the application objects they prove.

Best first task for an agent

Ask the agent to add a single event to one committed business action:

Add Lodger audit evidence for organization member removal.
Use the TypeScript SDK server-side.
Use action organization.member.removed.
Include actor, target, organizationId, requestId, sessionId, and the returned receipt ID.
Do not expose the Lodger API key to the browser.

After one event works, ask the agent to extract a small event catalog and apply the same pattern to related actions.