Build with one routed API.
LineMod is a security gateway with agent behavior observability—guardrails at the edge, harness diagnostics across runs, and policy routing when you need it. This page is the documentation overview for integrating through that gateway. Keep the OpenAI chat completions shape. LineMod routes, protects, and traces every request through gateway policy — with platform credits or BYOK on your own provider keys.
{
"model": "auto:balanced",
"messages": [
{ "role": "user", "content": "Summarize this incident log." }
]
}Three domains, one gateway hot path.
Use auto:* presets, org rules, failover chains, and cost-quality dials. Inspect
X-LineMod-Model-Used and X-LineMod-Route-Reason on every response.
Scan prompts and outputs, mask sensitive data, govern agent tool calls, and pin templates per API key. Blocked requests never reach your provider.
Per-request IDs, cost headers, agent run waterfalls, Agent insights with harness diagnostics, and exportable audit rows in the Console.
Documentation sections
auto:balanced, and copy-paste samples in Python, Node.js, Go, and cURL.https://api.linemod.ai/v1 or copy snippets from the Console.Platform credits or BYOK
X-LineMod-Credential-Source.| Mode | Who pays upstream | What you pay LineMod |
|---|---|---|
platform | LineMod wallet (platform credits) | Per-token usage bundled with routing, security, and observability |
org_byok | Your provider contract | Monthly BYOK subscription for gateway features |
platform_fallback | LineMod when BYOK path is unavailable (if enabled) | Fallback token usage per org policy |
Connect provider keys for OpenAI and Anthropic, route everything else through platform credits — one API key, automatic split billing. See pricing for plan details.