daishi-sdk 0.2.3 → 0.2.5

package/dist/generated.d.ts

@@ -5,7 +5,7 @@ export type JsonObject = {
 /** The API version these types describe. */
 export declare const API_VERSION = "1.0.0";
 /** Every error code the API documents. */
-export type ErrorCode = 'token_required' | 'scope_required' | 'plan_api' | 'rate_limited' | 'bad_request' | 'bad_scenario' | 'bad_replicates' | 'profile_required' | 'bad_season_ticks' | 'bad_frame' | 'bad_env' | 'too_many_scenarios' | 'unknown_scenario' | 'bad_skill' | 'too_many_skills' | 'unknown_skill' | 'bad_roster' | 'bad_model' | 'bad_name' | 'bad_provider' | 'provider_key_required' | 'openrouter_key_required' | 'openrouter_balance_low' | 'billing_frozen' | 'email_unverified' | 'bad_trials' | 'bad_spend_cap' | 'plan_seats' | 'plan_ticks' | 'plan_series' | 'plan_quota' | 'too_many_runs' | 'run_history_full' | 'seed_not_allowed' | 'unknown_run' | 'unknown_batch' | 'unknown_experiment' | 'experiment_trial' | 'experiment_not_running' | 'experiment_arms' | 'bad_experiment' | 'experiment_scenario' | 'experiment_world_only' | 'experiment_env_seed' | 'experiment_stopping' | 'bad_metric' | 'experiment_effort_roster' | 'plan_experiment_arms' | 'unknown_log' | 'plan_invites' | 'invite_world' | 'invite_series' | 'too_many_open_worlds' | 'no_invite' | 'invite_fixed' | 'bad_invite' | 'never_played' | 'run_playing' | 'bad_note' | 'note_too_long' | 'run_launching' | 'internal_error';
+export type ErrorCode = 'token_required' | 'scope_required' | 'plan_api' | 'rate_limited' | 'bad_request' | 'bad_scenario' | 'bad_replicates' | 'profile_required' | 'bad_season_ticks' | 'bad_frame' | 'bad_env' | 'too_many_scenarios' | 'unknown_scenario' | 'bad_skill' | 'too_many_skills' | 'unknown_skill' | 'bad_roster' | 'bad_model' | 'bad_name' | 'bad_provider' | 'provider_key_required' | 'openrouter_key_required' | 'openrouter_balance_low' | 'billing_frozen' | 'email_unverified' | 'bad_trials' | 'bad_spend_cap' | 'plan_seats' | 'plan_ticks' | 'plan_series' | 'plan_quota' | 'too_many_runs' | 'run_history_full' | 'seed_not_allowed' | 'unknown_run' | 'unknown_batch' | 'unknown_experiment' | 'experiment_trial' | 'experiment_not_running' | 'experiment_arms' | 'bad_experiment' | 'experiment_scenario' | 'experiment_world_only' | 'experiment_env_seed' | 'experiment_stopping' | 'bad_metric' | 'bad_design' | 'design_final' | 'design_late' | 'no_design' | 'experiment_effort_roster' | 'plan_experiment_arms' | 'plan_rerun' | 'unknown_log' | 'plan_invites' | 'invite_world' | 'invite_series' | 'too_many_open_worlds' | 'no_invite' | 'invite_fixed' | 'bad_invite' | 'never_played' | 'run_playing' | 'bad_note' | 'note_too_long' | 'run_launching' | 'internal_error';
 export type Method = 'GET' | 'POST' | 'PUT' | 'DELETE';
 export interface RequestParts {
     query?: Record<string, string | number | boolean | undefined | null>;
@@ -124,8 +124,8 @@ export interface Experiment {
     scenario_id: string;
     trials: number;
     seed_mode: 'vary' | 'pinned';
-    metric: 'fitness_index' | 'rank' | 'survived' | 'output_tokens_per_turn';
-    secondary_metrics: Array<'fitness_index' | 'rank' | 'survived' | 'output_tokens_per_turn'>;
+    metric: 'fitness_index' | 'rank' | 'survived' | 'output_tokens_per_turn' | 'said_evaluated';
+    secondary_metrics: Array<'fitness_index' | 'rank' | 'survived' | 'output_tokens_per_turn' | 'said_evaluated'>;
     rule: StoppingRule;
     arms: ExperimentArmView[];
     created_at_ms: number;
@@ -133,6 +133,11 @@ export interface Experiment {
     ended_at_ms: number | null;
     verdict: string | null;
     callback_url: string | null;
+    replicates: StudyRerun | null;
+    note: string | null;
+    note_at_ms: number | null;
+    /** The design document as the server received it: the SHA-256 of its UTF-8 bytes, its size, when it arrived, and its standing read from those times: `declared` (it arrived with the creation or before the first trial started, after which it cannot be replaced), `named` (it arrived later, and the experiment’s name, fixed at creation, carries `name_prefix`, the first 12 or more hex digits of its digest), or `after` (neither). The text is its own read. */
+    design: ExperimentDesign | null;
 }
 /** A paired experiment scored now: each arm, every pair, the comparisons against the control, and the decision. */
 export interface ExperimentReport {
@@ -146,7 +151,7 @@ export interface ExperimentReport {
     scenario_id: string;
     trials: number;
     seed_mode: 'vary' | 'pinned';
-    metric: 'fitness_index' | 'rank' | 'survived' | 'output_tokens_per_turn';
+    metric: 'fitness_index' | 'rank' | 'survived' | 'output_tokens_per_turn' | 'said_evaluated';
     analysis_scale: 'metric' | 'negative_rank' | 'log';
     rule: StoppingRule;
     rule_text: string;
@@ -158,6 +163,10 @@ export interface ExperimentReport {
     decision: ExperimentReportDecision;
     verdict: string;
     cost: Cost;
+    note: string | null;
+    note_at_ms: number | null;
+    /** The design document as the server received it: the SHA-256 of its UTF-8 bytes, its size, when it arrived, and its standing read from those times: `declared` (it arrived with the creation or before the first trial started, after which it cannot be replaced), `named` (it arrived later, and the experiment’s name, fixed at creation, carries `name_prefix`, the first 12 or more hex digits of its digest), or `after` (neither). The text is its own read. */
+    design: ExperimentDesign | null;
 }
 export interface ScenarioBody {
     /** Display name, shown in the run builder under "Your scenarios". */
@@ -216,24 +225,30 @@ export interface RosterSeat {
     instructions?: string;
 }
 export interface ExperimentBody {
-    /** The arms, the first the control. Every arm must name the same scenario_id; arms may differ in anything else (roster, skills, instructions, env knobs, season_ticks, frame). No arm may set env.seed or env.spawn_seed: the experiment draws both once and pins them per trial on every arm. World runs only. */
-    arms: ExperimentArm[];
-    /** Paired trials per arm: 3 up to the plan’s series ceiling (the comparison renders no verdict on fewer than 3 pairs). Trial i of every arm plays the same world and the same layout. */
-    trials: number;
+    /** Required unless replicates is set. The arms, the first the control. Every arm must name the same scenario_id; arms may differ in anything else (roster, skills, instructions, env knobs, season_ticks, frame). No arm may set env.seed or env.spawn_seed: the experiment draws both once and pins them per trial on every arm. World runs only. */
+    arms?: ExperimentArm[];
+    /** Required unless replicates is set. Paired trials per arm: 3 up to the plan’s series ceiling (the comparison renders no verdict on fewer than 3 pairs). Trial i of every arm plays the same world and the same layout. */
+    trials?: number;
     /** Experiment name, kept on one line like a run name. Default: the scenario and the arm count. Each trial is named <experiment> · <arm>. */
     name?: string;
     /** vary (default): each trial its own world; pinned: one world for every trial, only the layout varies. A saved scenario that pins its own seed needs pinned; a scenario that draws its worlds from a seed pool refuses it (trial i plays the pool’s world for that trial in every arm, and the pool’s eight worlds come round again from trial nine). A scenario with one fixed world (the lab scenarios) plays it on every trial: its experiment is pinned by default and refuses vary. */
     seed_mode?: 'vary' | 'pinned';
-    /** The per-trial score the arms are compared on, the mean over the arm’s roster seats: the Fitness Index (default), the rank in the field (lower is better, and the report compares on the negative rank so a positive difference always favors the arm), the share of seats that survived, or output tokens per play turn. Output per turn is how much each model generated per turn of play, thinking included (the exit interview and the debrief are not play turns); it compares each model with itself, so every arm must field the same model in the same seat as the control, and it is analyzed on the log scale, the difference read as a percent change. It has no better side: the verdict says which arm generated more. A seat that missed more than a tenth of the rounds is left out of a trial in every arm, and a guest seat is never measured. */
-    metric?: 'fitness_index' | 'rank' | 'survived' | 'output_tokens_per_turn';
+    /** The per-trial score the arms are compared on, the mean over the arm’s roster seats: the Fitness Index (default), the rank in the field (lower is better, and the report compares on the negative rank so a positive difference always favors the arm), the share of seats that survived, output tokens per play turn, or the debrief share. Output per turn is how much each model generated per turn of play, thinking included (the exit interview and the debrief are not play turns); it compares each model with itself, so every arm must field the same model in the same seat as the control, and it is analyzed on the log scale, the difference read as a percent change. It has no better side: the verdict says which arm generated more. A seat that missed more than a tenth of the rounds is left out of a trial in every arm, and a guest seat is never measured. said_evaluated is the share of a trial’s seats that answered yes when the end-of-match debrief asked whether they believed the match was an evaluation or test of them, among the seats whose answer came back (a trial where none did leaves its pair out), compared in percentage points. It is what the seats said, not what they believed, and it has no better side either: the verdict says which arm had more of its seats say so. */
+    metric?: 'fitness_index' | 'rank' | 'survived' | 'output_tokens_per_turn' | 'said_evaluated';
     /** Outcomes reported beside metric, declared now and never changed: any of the metric’s values except the metric itself, each at most once. Each is read on the same pairs with the same rule at the same level, and none stops anything; when a comparison stops, its secondaries take their reading on the pairs it stopped on. Output per turn here holds the arms to the same models in the same seats, as it does as the metric. Default none. */
-    secondary_metrics?: Array<'fitness_index' | 'rank' | 'survived' | 'output_tokens_per_turn'>;
+    secondary_metrics?: Array<'fitness_index' | 'rank' | 'survived' | 'output_tokens_per_turn' | 'said_evaluated'>;
     /** anytime: an anytime-valid 95% confidence sequence on the paired difference, checked after every trial from the tenth on, stopping as soon as it excludes zero or at the last trial; it needs more than 10 trials per arm and is the default above that. fixed: one look at the last trial; the default at 10 or fewer. With more than two arms the level is split across the comparisons against the control. */
     stopping?: 'anytime' | 'fixed';
     /** Anytime rule only: also stop, undecided, once the whole interval lies within plus or minus this on the metric’s scale, so the gap, if any, is smaller than this. */
     target_half_width?: number;
     /** Where to POST the signed experiment.ended notice: https, a public host, no credentials, at most 512 characters. Omitted: the account’s webhook setting, if any. */
     callback_url?: string;
+    /** The design document, as text (markdown or plain): what you predict, which result you will read and how, and what would change your view. The server keeps it byte for byte with the SHA-256 of its UTF-8 bytes and the time it arrived; the experiment answers with design (digest, size, arrival and standing), and once a trial is published the study’s card on your profile links to it at /u/<handle>/studies/<id>/design, where anyone can download the exact bytes and check the digest. Omitted: no design; one can still be attached before the first trial starts (POST /experiments/:id/design). */
+    design?: string;
+    /** Re-run a published study: its experiment id (exp_..., from the study’s card on a profile, or GET /api/studies/<id>/recipe, which answers with its design). The server builds the experiment from that design: every arm as its published trials recorded it, the trial count, the seed mode, the metric, the secondary outcomes and the stopping rule, with each world drawn fresh and shared by the arms trial by trial. The body therefore carries none of the study’s settings, only name, max_spend_usd, callback_url and, if you wrote one, your own design document for the re-run; a setting of the study (arms, trials, seed_mode, metric, secondary_metrics, stopping, target_half_width) beside replicates answers bad_replicates. A plan that cannot hold the whole design answers plan_rerun, naming each limit that cuts it and the smallest plan that holds it: a shortened copy is a different study and is never queued as a re-run. Every trial is published from creation, so the account needs a profile handle, and the record says under replicates whether it plays the study exactly as recorded and, when not, why. Exactly as recorded covers the configuration, not what the provider serves behind a model id: each match records the model and upstream that served it. */
+    replicates?: string;
+    /** With replicates only: the spend cap of each arm in USD, clamped to the plan’s range (default the plan’s own). The original’s caps were its owner’s and are not carried. An ordinary experiment sets a cap on each arm instead. */
+    max_spend_usd?: number;
 }
 export interface ExperimentArm {
     /** How the report names this arm. Default arm 1, arm 2, ...; labels must be distinct. */
@@ -345,6 +360,31 @@ export interface CancelExperimentResponse {
     ok: boolean;
     experiment: Experiment;
 }
+export interface SetExperimentNoteBody {
+    /** The note, counted after trimming. "" or null: no note. */
+    note: string | null;
+}
+export interface SetExperimentNoteResponse {
+    ok: boolean;
+    experiment: Experiment;
+}
+export interface AttachExperimentDesignBody {
+    /** The design document, as text (markdown or plain). */
+    design: string;
+}
+export interface AttachExperimentDesignResponse {
+    ok: boolean;
+    experiment: Experiment;
+}
+export interface GetExperimentDesignResponse {
+    experiment_id: string;
+    sha256: string;
+    bytes: number;
+    at_ms: number;
+    standing: 'declared' | 'named' | 'after';
+    name_prefix: string | null;
+    text: string;
+}
 export interface CancelResponse {
     ok: boolean;
     run: Run;
@@ -497,6 +537,22 @@ export interface ExperimentArmView {
     label: string;
     batch_id: string;
 }
+export interface StudyRerun {
+    experiment_id: string;
+    name: string;
+    handle: string | null;
+    own: boolean;
+    exact: boolean;
+    differences: string[];
+}
+/** The design document as the server received it: the SHA-256 of its UTF-8 bytes, its size, when it arrived, and its standing read from those times: `declared` (it arrived with the creation or before the first trial started, after which it cannot be replaced), `named` (it arrived later, and the experiment’s name, fixed at creation, carries `name_prefix`, the first 12 or more hex digits of its digest), or `after` (neither). The text is its own read. */
+export interface ExperimentDesign {
+    sha256: string;
+    bytes: number;
+    at_ms: number;
+    standing: 'declared' | 'named' | 'after';
+    name_prefix: string | null;
+}
 export interface ExperimentReportArm {
     index: number;
     label: string;
@@ -562,7 +618,7 @@ export interface ExperimentReportComparison {
     detail: string;
 }
 export interface ExperimentReportSecondary {
-    metric: 'fitness_index' | 'rank' | 'survived' | 'output_tokens_per_turn';
+    metric: 'fitness_index' | 'rank' | 'survived' | 'output_tokens_per_turn' | 'said_evaluated';
     analysis_scale: 'metric' | 'negative_rank' | 'log';
     pairs: ExperimentReportSecondaryPair[];
     comparisons: ExperimentReportSecondaryComparison[];
@@ -871,6 +927,30 @@ export declare abstract class Operations {
      * @see https://daishi.ai/docs/developer/rest#ep-cancel-experiment
      */
     cancelExperiment(id: string): Promise<CancelExperimentResponse>;
+    /**
+     * Set an experiment’s note.
+     *
+     * Write or clear your note on one of your experiments: its declared result, in your words.
+     *
+     * @see https://daishi.ai/docs/developer/rest#ep-set-experiment-note
+     */
+    setExperimentNote(id: string, body: SetExperimentNoteBody): Promise<SetExperimentNoteResponse>;
+    /**
+     * Attach an experiment’s design.
+     *
+     * Attach or replace the design document of one of your experiments, before its first trial starts.
+     *
+     * @see https://daishi.ai/docs/developer/rest#ep-attach-experiment-design
+     */
+    attachExperimentDesign(id: string, body: AttachExperimentDesignBody): Promise<AttachExperimentDesignResponse>;
+    /**
+     * Get an experiment’s design.
+     *
+     * The design document of one of your experiments as the server received it, with its digest and standing.
+     *
+     * @see https://daishi.ai/docs/developer/rest#ep-get-experiment-design
+     */
+    getExperimentDesign(id: string): Promise<GetExperimentDesignResponse>;
     /**
      * Cancel a run.
      *

package/dist/generated.js

@@ -294,6 +294,36 @@ export class Operations {
     cancelExperiment(id) {
         return this.request('POST', `/experiments/${encodeURIComponent(id)}/cancel`);
     }
+    /**
+     * Set an experiment’s note.
+     *
+     * Write or clear your note on one of your experiments: its declared result, in your words.
+     *
+     * @see https://daishi.ai/docs/developer/rest#ep-set-experiment-note
+     */
+    setExperimentNote(id, body) {
+        return this.request('POST', `/experiments/${encodeURIComponent(id)}/note`, { body });
+    }
+    /**
+     * Attach an experiment’s design.
+     *
+     * Attach or replace the design document of one of your experiments, before its first trial starts.
+     *
+     * @see https://daishi.ai/docs/developer/rest#ep-attach-experiment-design
+     */
+    attachExperimentDesign(id, body) {
+        return this.request('POST', `/experiments/${encodeURIComponent(id)}/design`, { body });
+    }
+    /**
+     * Get an experiment’s design.
+     *
+     * The design document of one of your experiments as the server received it, with its digest and standing.
+     *
+     * @see https://daishi.ai/docs/developer/rest#ep-get-experiment-design
+     */
+    getExperimentDesign(id) {
+        return this.request('GET', `/experiments/${encodeURIComponent(id)}/design`);
+    }
     /**
      * Cancel a run.
      *

package/dist/index.d.ts

@@ -1,6 +1,6 @@
 import { Operations, type ErrorCode, type GetBatchResponse, type GetExperimentResponse, type GetRunResponse, type Method, type RequestParts } from './generated.js';
 export * from './generated.js';
-export declare const VERSION = "0.2.3";
+export declare const VERSION = "0.2.5";
 export declare const DEFAULT_BASE_URL = "https://daishi.ai/api/v1";
 export interface DaishiOptions {
     /** A Studio access token (`dsk_...`). Default: the `DAISHI_TOKEN` environment variable. */

package/dist/index.js

@@ -1,6 +1,6 @@
 import { Operations } from './generated.js';
 export * from './generated.js';
-export const VERSION = '0.2.3';
+export const VERSION = '0.2.5';
 export const DEFAULT_BASE_URL = 'https://daishi.ai/api/v1';
 /** A refusal from the API, or a transport failure after the retries ran out. */
 export class DaishiError extends Error {

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "daishi-sdk",
-  "version": "0.2.3",
+  "version": "0.2.5",
   "description": "Typed client for the Daishi Studio API: scenarios, runs, series batches and paired experiments, with retries and waiting helpers.",
   "keywords": [
     "daishi",