amosblomqvist
f9e3a4c43d
Model slots wave (2026-09-20), chunk B:
- settings: new `models` section {strong, cheap, vision} slots (each
field independent, same shape as a role override); ROLE_SLOT maps all
nine roles to tiers (strong=prober+teacher, cheap=researcher/
demoBuilder/extractor/auditor/labeler, vision=visualMaker+pdfReader);
resolution chain per role: override → slot → factory pin.
- config: ModelSpec gains provider (all pins openrouter, the BYOK
default); MODEL_PINS in stores.ts now spans all nine roles (the strong
slot's factory rep IS the prober pin).
- engine: piModels() registers EVERY builtin provider catalog;
piEngineForSpec(resolved spec) is the new engine factory (pi-ai
resolves per-provider keys from env itself); piEngineFor stays as the
config-pin fallback. chat-handler resolves the conversation chain
(chat override → strong slot → pin) and injects engineFor into the job
runner, closing the old gap where per-role overrides never reached
job engines.
- /api/models: models federate ONLY keyed providers (providers surface
with keyPresent, env-wins, fresh per call); slots (resolved tier pins)
join rolePins; catalogDefaultsFor inherits the strong slot; the models
settings section gets the same catalog gate as chat.
- pricing: PRICES/CONTEXT_LIMITS uniformly keyed provider:id
(regenerated from ALL pi-ai catalogs, 1354 keys); usage stamps
(TurnUsage, JobUsage, extraction stages) switch to provider:id —
owner decision: no legacy compat, old stamps render unpriced.
77 lines
3.3 KiB
TypeScript
77 lines
3.3 KiB
TypeScript
/**
|
|
* Model-catalog contracts (model-control wave, 2026-09-12) — the client-safe
|
|
* shapes of GET /api/models and PUT /api/chats/[id]/settings responses.
|
|
* Pure types; the server projection lives in src/server/model-catalog.ts
|
|
* (which OWNS these shapes — the contract is defined here only so client
|
|
* components never import from the server band).
|
|
*/
|
|
import type { ThinkingLevel } from "@earendil-works/pi-agent-core";
|
|
import type { BackgroundModelRole, ModelSlotName } from "../settings.ts";
|
|
|
|
/** One picker row: everything the UI shows about a model, projected from
|
|
* pi-ai's catalog. Costs are USD per 1M tokens (the same catalog, the same
|
|
* units core/prices.ts's generated table carries). */
|
|
export interface ModelCatalogEntry {
|
|
id: string;
|
|
name: string;
|
|
provider: string;
|
|
reasoning: boolean;
|
|
input: ("text" | "image")[];
|
|
/** The thinking levels this model actually accepts (pi's
|
|
* thinkingLevelMap — e.g. GLM-5.3-flash: low|high|max). The thinking
|
|
* picker lists exactly these; anything else is a lie. */
|
|
supportedLevels: ThinkingLevel[];
|
|
cost: { input: number; output: number; cacheRead: number; cacheWrite: number };
|
|
contextWindow: number;
|
|
maxTokens: number;
|
|
}
|
|
|
|
/** The conversation agents' config pin (config.ts MODELS prober/teacher —
|
|
* one controller, so one pin): what a chat without settings runs. */
|
|
export interface CatalogDefaults {
|
|
model: { provider: string; id: string };
|
|
thinkingLevel: ThinkingLevel;
|
|
}
|
|
|
|
/** One background role's RESOLVED pin (settings interlude, 2026-09-19):
|
|
* the user's agent-models override falling back to the config.ts MODELS
|
|
* pin — what that role actually runs today. The settings UI shows each
|
|
* agent-role row's "Default: <this>" caption from it. */
|
|
export type RolePins = Record<
|
|
BackgroundModelRole,
|
|
{ model: { provider: string; id: string }; thinkingLevel?: ThinkingLevel }
|
|
>;
|
|
|
|
/** One slot tier's RESOLVED pin (model slots wave, 2026-09-20): the
|
|
* strong/cheap/vision tier's concrete model — slot override falling back
|
|
* to the factory rep pin. The settings UI captions each slot row with it
|
|
* and the picker pre-fills from it. */
|
|
export type SlotPins = Record<ModelSlotName, { model: { provider: string; id: string }; thinkingLevel?: ThinkingLevel }>;
|
|
|
|
/** GET /api/models response. */
|
|
export interface ModelCatalogResponse {
|
|
models: ModelCatalogEntry[];
|
|
defaults: CatalogDefaults;
|
|
/** The six background roles' resolved pins (same resolution the agent
|
|
* factories use — src/server/stores.ts effectiveSettings). */
|
|
rolePins: RolePins;
|
|
/** The three slot tiers' resolved pins (model slots wave, 2026-09-20):
|
|
* same resolution chain — role override → slot → factory pin — reported
|
|
* at tier granularity; effectiveSettings().models verbatim. */
|
|
slots: SlotPins;
|
|
/** Providers with a PRESENT key, and which env var gates each — the
|
|
* model slots wave's federation surface (2026-09-20): the picker lists
|
|
* models from these providers' pi-ai catalogs only. Keyless providers
|
|
* are simply absent (the Keys page is the discovery surface, same
|
|
* discipline as the search engines). */
|
|
providers: CatalogProvider[];
|
|
}
|
|
|
|
/** One row of the providers surface. */
|
|
export interface CatalogProvider {
|
|
id: string;
|
|
/** The env var whose presence federates this provider in. */
|
|
envVar: string;
|
|
/** Whether that var is currently present (env or auth file). */
|
|
keyPresent: boolean;
|
|
} |