Skip to content

Self-hosted, Self-Improving, Federated Everything App — dev management, agent orchestration, persistent mind, digital twin, second brain, and creative pipeline from a single dashboard over Tailscale to port the user into the OS

1.6 KB
Raw Blame
// Shared resolution for "which provider/model should a Pipeline LLM action
// use?" — falls back to the series' configured LLM when the client doesn't
// pass an explicit override, so every Pipeline action (extract-canon,
// extract-scenes, season-episodes-generate) honors the provider/model picked
// in the series header instead of the global active provider.
//
// The never-cross-providers rule this rests on lives in `llmRoutePin.js`: a
// model id belongs to the provider it was picked for, so the series model is
// only inherited while the EFFECTIVE provider is still the series provider.
// This module is the series-shaped adapter over it — `series.llm` spells the
// pin `{ provider, model }` and callers want `undefined` (not `null`) for an
// unresolved field so the result passes straight through to the extractor.

import { resolveLlmRoutePin } from './llmRoutePin.js';

/**
 * Resolve the effective LLM provider/model for a Pipeline action against a series.
 *
 * @param {{ llm?: { provider?: string, model?: string } } | null | undefined} series
 * @param {{ overrideProvider?: string, overrideModel?: string }} [overrides]
 * @returns {{ provider: string|undefined, model: string|undefined, providerMatchesSeries: boolean }}
 */
export function resolveSeriesLlmOverride(series, { overrideProvider, overrideModel } = {}) {
  const { providerId, model, providerMatchesPin } = resolveLlmRoutePin(
    { providerId: series?.llm?.provider, model: series?.llm?.model },
    { providerId: overrideProvider, model: overrideModel },
  );
  return {
    provider: providerId ?? undefined,
    model: model ?? undefined,
    providerMatchesSeries: providerMatchesPin,
  };
}