jev-recipes 0.5.2 → 0.5.3

package/README.md

@@ -11,7 +11,7 @@
 
 <!-- BEGIN GENERATED: summary -->
 
-87 focused recipes for JavaScript and TypeScript. Route messages, check evidence, and label model responses with a function call.
+88 focused recipes for JavaScript and TypeScript. Route messages, check evidence, and label model responses with a function call.
 
 <!-- END GENERATED: summary -->
 
@@ -138,7 +138,7 @@ npx jev-recipes demo answerability
 
 [Browse the complete catalog](https://github.com/agencyenterprise/jev-recipes/blob/main/recipes/README.md). Each guide includes an import, input reference, result behavior, limitations, and related recipes.
 
-For games, the [gameplay guide](https://github.com/agencyenterprise/jev-recipes/blob/main/docs/gameplay.md) starts with a checkers decision using just the board, player, and legal moves. It also covers general action selection and narrative turn eligibility.
+For games, use [game-action](https://github.com/agencyenterprise/jev-recipes/blob/main/recipes/game-action/README.md) with your existing JSON state and actions. It returns the original selected action, preserving your game IDs. [checkers-move](https://github.com/agencyenterprise/jev-recipes/blob/main/recipes/checkers-move/README.md) accepts a structured checkers board and legal moves; the [checkers example](https://github.com/agencyenterprise/jev-recipes/blob/main/examples/checkers/README.md) shows how to run its visual demo locally.
 
 For psychology, browse [Psychology & behavior](https://github.com/agencyenterprise/jev-recipes/blob/main/recipes/README.md#psychology--behavior) for gain/loss framing, causal explanations, stated motivation, and related wording annotations. Search with `npx jev-recipes list psychology`.
 
@@ -185,7 +185,7 @@ The [CI workflow](https://github.com/agencyenterprise/jev-recipes/actions/workfl
 - **Recipe checks:** validate software behavior using mocked responses, including invalid inputs, confidence boundaries, and malformed answers.
 - **Model evaluation:** offline tests and demos do not measure Jev's accuracy. Evaluate recipes on your own data for your intended use.
 
-The package contains compiled code, type declarations, catalog data, and offline demo assets. Tests, build tools, and source guides stay out of the npm archive. [Publishing details](https://github.com/agencyenterprise/jev-recipes/blob/main/RELEASING.md).
+The package contains compiled code, type declarations, catalog data, and offline demo assets. Tests, build tools, and source guides stay out of the npm archive. [Publishing details](https://github.com/agencyenterprise/jev-recipes/blob/main/CONTRIBUTING.md#releasing).
 
 Direct imports such as `jev-recipes/route` load the selected recipe and its dependencies. Installation downloads one package; it does not selectively download individual recipes.
 
@@ -193,7 +193,7 @@ Direct imports such as `jev-recipes/route` load the selected recipe and its depe
 
 Recipes can serve as candidate annotation tools in controlled experiments. `claim-stance` labels expressed agreement, `verify` checks supplied evidence, and `draft-compare` compares responses under a rubric.
 
-The [research guide](https://github.com/agencyenterprise/jev-recipes/blob/main/docs/ai-alignment-research.md) outlines a controlled agreement study and validation against human annotations. These labels describe observable outputs; they do not establish internal motives or prove that a model is aligned.
+The [research guide](https://github.com/agencyenterprise/jev-recipes/blob/main/docs/ai-alignment-research.md) covers validation against human annotations, handling uncertainty, and recording reproducible results. These labels describe observable outputs; they do not establish internal motives or prove that a model is aligned.
 
 <a id="contribute"></a>
 
@@ -211,7 +211,7 @@ make ci
 
 The recipe folder owns its implementation, schemas, metadata, demo, and guide. Exports and catalog entries are generated from those folders. Tests live under `tests/recipe/`.
 
-[Contributing guide](https://github.com/agencyenterprise/jev-recipes/blob/main/CONTRIBUTING.md) | [Roadmap](https://github.com/agencyenterprise/jev-recipes/blob/main/RECIPE_ROADMAP.md) | [Changelog](https://github.com/agencyenterprise/jev-recipes/blob/main/CHANGELOG.md)
+[Contributing guide](https://github.com/agencyenterprise/jev-recipes/blob/main/CONTRIBUTING.md)
 
 ## License
 

package/dist/catalog/generated/details/checkers-move.json

@@ -33,7 +33,6 @@
       },
       "legalMoves": {
         "minItems": 1,
-        "maxItems": 50,
         "type": "array",
         "items": {
           "type": "object",
@@ -72,7 +71,7 @@
           "required": ["id", "from", "path"],
           "additionalProperties": false
         },
-        "description": "One to fifty complete legal moves from your game, with unique IDs."
+        "description": "One or more complete legal moves from your game, with unique IDs."
       },
       "minConfidence": { "type": "number", "minimum": 0, "maximum": 1 }
     },

package/dist/catalog/generated/details/choose-action.json

@@ -21,14 +21,13 @@
       },
       "actions": {
         "minItems": 1,
-        "maxItems": 50,
         "type": "array",
         "items": {
           "type": "object",
           "properties": { "id": { "type": "string" }, "text": { "type": "string" } },
           "required": ["id", "text"]
         },
-        "description": "One to fifty candidate actions with unique IDs and concrete action descriptions."
+        "description": "One or more candidate actions with unique IDs and concrete action descriptions."
       },
       "history": {
         "maxItems": 100,

package/dist/catalog/generated/details/game-action.json

@@ -0,0 +1,161 @@
+{
+  "inputSchema": {
+    "$schema": "https://json-schema.org/draft/2020-12/schema",
+    "type": "object",
+    "properties": {
+      "gameState": {
+        "description": "The current game information available for this decision.",
+        "$ref": "#/$defs/__schema0"
+      },
+      "playerState": {
+        "description": "Information about the player making this decision.",
+        "$ref": "#/$defs/__schema1"
+      },
+      "legalActions": {
+        "type": "array",
+        "items": { "$ref": "#/$defs/__schema2" },
+        "description": "The available actions the player can choose from."
+      },
+      "rules": {
+        "description": "Game rules and explanations of game-specific information.",
+        "$ref": "#/$defs/__schema3"
+      },
+      "objective": {
+        "description": "What the player is trying to accomplish.",
+        "$ref": "#/$defs/__schema4"
+      }
+    },
+    "required": ["gameState", "playerState", "legalActions"],
+    "$defs": {
+      "__schema0": {
+        "anyOf": [
+          { "type": "string" },
+          { "type": "number" },
+          { "type": "boolean" },
+          { "type": "null" },
+          { "type": "array", "items": { "$ref": "#/$defs/__schema0" } },
+          {
+            "type": "object",
+            "propertyNames": { "type": "string" },
+            "additionalProperties": { "$ref": "#/$defs/__schema0" }
+          }
+        ]
+      },
+      "__schema1": {
+        "anyOf": [
+          { "type": "string" },
+          { "type": "number" },
+          { "type": "boolean" },
+          { "type": "null" },
+          { "type": "array", "items": { "$ref": "#/$defs/__schema1" } },
+          {
+            "type": "object",
+            "propertyNames": { "type": "string" },
+            "additionalProperties": { "$ref": "#/$defs/__schema1" }
+          }
+        ]
+      },
+      "__schema2": {
+        "anyOf": [
+          { "type": "string" },
+          { "type": "number" },
+          { "type": "boolean" },
+          { "type": "null" },
+          { "type": "array", "items": { "$ref": "#/$defs/__schema2" } },
+          {
+            "type": "object",
+            "propertyNames": { "type": "string" },
+            "additionalProperties": { "$ref": "#/$defs/__schema2" }
+          }
+        ]
+      },
+      "__schema3": {
+        "anyOf": [
+          { "type": "string" },
+          { "type": "number" },
+          { "type": "boolean" },
+          { "type": "null" },
+          { "type": "array", "items": { "$ref": "#/$defs/__schema3" } },
+          {
+            "type": "object",
+            "propertyNames": { "type": "string" },
+            "additionalProperties": { "$ref": "#/$defs/__schema3" }
+          }
+        ]
+      },
+      "__schema4": {
+        "anyOf": [
+          { "type": "string" },
+          { "type": "number" },
+          { "type": "boolean" },
+          { "type": "null" },
+          { "type": "array", "items": { "$ref": "#/$defs/__schema4" } },
+          {
+            "type": "object",
+            "propertyNames": { "type": "string" },
+            "additionalProperties": { "$ref": "#/$defs/__schema4" }
+          }
+        ]
+      }
+    }
+  },
+  "resultSchema": {
+    "$schema": "https://json-schema.org/draft/2020-12/schema",
+    "anyOf": [
+      {
+        "type": "object",
+        "properties": {
+          "model": { "type": "string" },
+          "usage": {
+            "type": "object",
+            "properties": {
+              "input_tokens": { "type": "integer", "minimum": 0, "maximum": 9007199254740991 },
+              "output_tokens": { "type": "integer", "minimum": 0, "maximum": 9007199254740991 }
+            },
+            "required": ["input_tokens", "output_tokens"],
+            "additionalProperties": false
+          },
+          "selection": {
+            "type": "string",
+            "description": "The internal Choice label associated with the selected action."
+          },
+          "action": {
+            "description": "The original selected action, preserving any existing game ID.",
+            "$ref": "#/$defs/__schema0"
+          },
+          "confidence": {
+            "type": "number",
+            "minimum": 0,
+            "maximum": 1,
+            "description": "The confidence returned by Jev."
+          },
+          "probabilities": {
+            "type": "object",
+            "propertyNames": { "type": "string" },
+            "additionalProperties": { "type": "number", "minimum": 0, "maximum": 1 },
+            "description": "Jev's probabilities, keyed by internal Choice label."
+          }
+        },
+        "required": ["model", "usage", "selection", "action", "confidence", "probabilities"],
+        "additionalProperties": false
+      },
+      { "type": "null" }
+    ],
+    "$defs": {
+      "__schema0": {
+        "anyOf": [
+          { "type": "string" },
+          { "type": "number" },
+          { "type": "boolean" },
+          { "type": "null" },
+          { "type": "array", "items": { "$ref": "#/$defs/__schema0" } },
+          {
+            "type": "object",
+            "propertyNames": { "type": "string" },
+            "additionalProperties": { "$ref": "#/$defs/__schema0" }
+          }
+        ]
+      }
+    }
+  }
+}

package/dist/catalog/generated/loaders.js

@@ -37,6 +37,7 @@ export const recipeLoaders = {
     'followup-link': () => import('../../recipes/followup-link/index.js').then((recipe) => (input, options) => recipe.followupLink(recipe.followupLinkInputSchema.parse(input), options)),
     'freshness-needed': () => import('../../recipes/freshness-needed/index.js').then((recipe) => (input, options) => recipe.freshnessNeeded(recipe.freshnessNeededInputSchema.parse(input), options)),
     'frustration-signal': () => import('../../recipes/frustration-signal/index.js').then((recipe) => (input, options) => recipe.frustrationSignal(recipe.frustrationSignalInputSchema.parse(input), options)),
+    'game-action': () => import('../../recipes/game-action/index.js').then((recipe) => (input, options) => recipe.gameAction(recipe.gameActionInputSchema.parse(input), options)),
     handoff: () => import('../../recipes/handoff/index.js').then((recipe) => (input, options) => recipe.handoff(recipe.handoffInputSchema.parse(input), options)),
     'incident-match': () => import('../../recipes/incident-match/index.js').then((recipe) => (input, options) => recipe.incidentMatch(recipe.incidentMatchInputSchema.parse(input), options)),
     'instruction-conflict': () => import('../../recipes/instruction-conflict/index.js').then((recipe) => (input, options) => recipe.instructionConflict(recipe.instructionConflictInputSchema.parse(input), options)),

package/dist/catalog/generated/metadata.js

@@ -784,6 +784,29 @@ export const recipeMetadata = [
             },
         ],
     },
+    {
+        id: 'game-action',
+        title: 'Choose a game action from JSON',
+        description: 'Send game state, player state, and available actions as JSON; receive the original selected action with Jev confidence and probabilities.',
+        category: 'workflow',
+        tags: ['gameplay', 'game', 'agnostic', 'JSON', 'actions', 'automatic IDs', 'simulation'],
+        limitations: [
+            'Confidence and probabilities describe the Jev decision, not a measured probability of winning.',
+            'The full action list is sent in one logical request. Provider request limits still apply.',
+            'Internal action labels belong to the current call; existing game IDs remain in the returned action.',
+        ],
+        useWhen: 'Your game already supplies JSON state and actions, and you want Jev to choose an action without formatting descriptions or assigning IDs.',
+        related: [
+            {
+                id: 'choose-action',
+                reason: 'Use choose-action for text-based rules and state, caller-named actions, and confidence-based review handling.',
+            },
+            {
+                id: 'checkers-move',
+                reason: 'Use checkers-move for its American/English checkers board format and built-in checkers instructions.',
+            },
+        ],
+    },
     {
         id: 'handoff',
         title: 'Check handoff rules',

package/dist/catalog/generated/names.d.ts

@@ -1 +1 @@
-export declare const recipeNames: readonly ["action-scope", "answer-consistency", "answer-coverage", "answer-invalidation", "answer-relevance", "answerability", "argument-fit", "attempted-step", "attribution-match", "audience-fit", "cache-match", "cancellation-check", "causal-attribution", "certainty-match", "change-meaning", "checkers-move", "choose-action", "citation-match", "citation-needed", "claim-stance", "clarify", "confirmation-match", "constraint-strength", "context-role", "correction-target", "document-role", "draft-compare", "evaluation-mention", "evidence-conflict", "evidence-independence", "evidence-novelty", "fact-stability", "failure-kind", "feedback-kind", "field-select", "followup-link", "freshness-needed", "frustration-signal", "handoff", "incident-match", "instruction-conflict", "instruction-fit", "intent-change", "issue-impact", "issue-recurrence", "memory-relation", "memory-scope", "memory-subject", "memory-value", "motivation-source", "outcome-framing", "passage-duplicate", "preference-kind", "promise-check", "query-equivalence", "query-specificity", "question-assumption", "question-leading", "reference-resolve", "repeated-attempt", "reply-template-match", "requirement-testability", "rerank", "resolution-check", "response-needed", "response-refusal", "result-outcome", "result-usefulness", "retrieval-needed", "route", "source-applicability", "step-complete", "step-progress", "summary-coverage", "take-turn", "task-dependency", "task-duplicate", "ticket-match", "tone-check", "tool-fit", "topic-shift", "troubleshooting-fit", "turn-intent", "uncertainty-expression", "urgency-signal", "verify", "workaround-fit"];
+export declare const recipeNames: readonly ["action-scope", "answer-consistency", "answer-coverage", "answer-invalidation", "answer-relevance", "answerability", "argument-fit", "attempted-step", "attribution-match", "audience-fit", "cache-match", "cancellation-check", "causal-attribution", "certainty-match", "change-meaning", "checkers-move", "choose-action", "citation-match", "citation-needed", "claim-stance", "clarify", "confirmation-match", "constraint-strength", "context-role", "correction-target", "document-role", "draft-compare", "evaluation-mention", "evidence-conflict", "evidence-independence", "evidence-novelty", "fact-stability", "failure-kind", "feedback-kind", "field-select", "followup-link", "freshness-needed", "frustration-signal", "game-action", "handoff", "incident-match", "instruction-conflict", "instruction-fit", "intent-change", "issue-impact", "issue-recurrence", "memory-relation", "memory-scope", "memory-subject", "memory-value", "motivation-source", "outcome-framing", "passage-duplicate", "preference-kind", "promise-check", "query-equivalence", "query-specificity", "question-assumption", "question-leading", "reference-resolve", "repeated-attempt", "reply-template-match", "requirement-testability", "rerank", "resolution-check", "response-needed", "response-refusal", "result-outcome", "result-usefulness", "retrieval-needed", "route", "source-applicability", "step-complete", "step-progress", "summary-coverage", "take-turn", "task-dependency", "task-duplicate", "ticket-match", "tone-check", "tool-fit", "topic-shift", "troubleshooting-fit", "turn-intent", "uncertainty-expression", "urgency-signal", "verify", "workaround-fit"];

package/dist/catalog/generated/names.js

@@ -38,6 +38,7 @@ export const recipeNames = [
     'followup-link',
     'freshness-needed',
     'frustration-signal',
+    'game-action',
     'handoff',
     'incident-match',
     'instruction-conflict',

package/dist/catalog/schema.d.ts

@@ -38,6 +38,7 @@ export declare const recipeNameSchema: z.ZodEnum<{
     "followup-link": "followup-link";
     "freshness-needed": "freshness-needed";
     "frustration-signal": "frustration-signal";
+    "game-action": "game-action";
     handoff: "handoff";
     "incident-match": "incident-match";
     "instruction-conflict": "instruction-conflict";

package/dist/cli/schema.d.ts

@@ -41,6 +41,7 @@ export declare const commandArgumentsSchema: z.ZodUnion<readonly [z.ZodTuple<[z.
     "followup-link": "followup-link";
     "freshness-needed": "freshness-needed";
     "frustration-signal": "frustration-signal";
+    "game-action": "game-action";
     handoff: "handoff";
     "incident-match": "incident-match";
     "instruction-conflict": "instruction-conflict";
@@ -132,6 +133,7 @@ export declare const commandArgumentsSchema: z.ZodUnion<readonly [z.ZodTuple<[z.
     "followup-link": "followup-link";
     "freshness-needed": "freshness-needed";
     "frustration-signal": "frustration-signal";
+    "game-action": "game-action";
     handoff: "handoff";
     "incident-match": "incident-match";
     "instruction-conflict": "instruction-conflict";
@@ -220,6 +222,7 @@ export declare const commandArgumentsSchema: z.ZodUnion<readonly [z.ZodTuple<[z.
     "followup-link": "followup-link";
     "freshness-needed": "freshness-needed";
     "frustration-signal": "frustration-signal";
+    "game-action": "game-action";
     handoff: "handoff";
     "incident-match": "incident-match";
     "instruction-conflict": "instruction-conflict";

package/dist/recipes/checkers-move/schema.js

@@ -43,9 +43,8 @@ export const checkersMoveInputSchema = z
     legalMoves: z
         .array(move)
         .min(1)
-        .max(50)
         .refine((moves) => new Set(moves.map((value) => value.id)).size === moves.length, 'Move IDs must be unique.')
-        .describe('One to fifty complete legal moves from your game, with unique IDs.'),
+        .describe('One or more complete legal moves from your game, with unique IDs.'),
     minConfidence: probability.optional(),
 })
     .superRefine((input, context) => {

package/dist/recipes/choose-action/schema.js

@@ -1,11 +1,15 @@
 import { z } from 'zod';
-import { nonEmptyText, probability, selectionResultSchema, textItemsSchema, } from '../../src/schema.js';
+import { nonEmptyText, probability, selectionResultSchema, textItemSchema, } from '../../src/schema.js';
 export const chooseActionInputSchema = z.object({
     player: nonEmptyText.describe('The player whose next action should be chosen.'),
     objective: nonEmptyText.describe('The goal or scoring preference used to compare eligible actions.'),
     rules: nonEmptyText.describe('Applicable game rules, including turn, phase, resource, and action constraints.'),
     environment: nonEmptyText.describe('The current game snapshot available to the player, after the supplied history.'),
-    actions: textItemsSchema.describe('One to fifty candidate actions with unique IDs and concrete action descriptions.'),
+    actions: z
+        .array(textItemSchema)
+        .min(1)
+        .refine((actions) => new Set(actions.map((action) => action.id)).size === actions.length, 'Action IDs must be unique.')
+        .describe('One or more candidate actions with unique IDs and concrete action descriptions.'),
     history: z
         .array(z.object({
         player: nonEmptyText.describe('The player who took the recorded action.'),

package/dist/recipes/game-action/demo.json

@@ -0,0 +1,24 @@
+{
+  "input": {
+    "gameState": { "actionsRemaining": 1, "opponentScore": 9 },
+    "playerState": { "id": "blue", "score": 8, "energy": 1 },
+    "legalActions": [
+      { "id": "collect-points", "command": "collect", "energyCost": 1, "points": 2 },
+      { "command": "rest", "energyGain": 1, "points": 0 }
+    ],
+    "rules": "The game ends after this action. Highest score wins; unused energy has no value.",
+    "objective": "Finish with more points than the opponent."
+  },
+  "response": {
+    "model": "demo-fixture",
+    "answers": {
+      "decision": {
+        "type": "choice",
+        "choice": "action_0",
+        "confidence": 0.9,
+        "probabilities": { "action_0": 0.95, "action_1": 0.05 }
+      }
+    },
+    "usage": { "input_tokens": 0, "output_tokens": 0 }
+  }
+}

package/dist/recipes/game-action/index.d.ts

@@ -0,0 +1,5 @@
+import type { RecipeOptions } from '../../src/schema.js';
+import type { GameActionInput, GameActionResult } from './schema.js';
+export declare function gameAction(input: GameActionInput, options?: RecipeOptions): Promise<GameActionResult>;
+export { gameActionInputSchema, gameActionResultSchema } from './schema.js';
+export type { GameActionInput, GameActionResult } from './schema.js';

package/dist/recipes/game-action/index.js

@@ -0,0 +1,30 @@
+import { choice } from '@typesafe-ai/sdk';
+import { evaluateWithJev } from '../../src/client.js';
+import { parseChoiceAnswer } from '../../src/answers.js';
+import { parseDecisionState } from '../../src/decisions.js';
+import { gameActionInputSchema, gameActionResultSchema } from './schema.js';
+const instructions = [
+    'Which available action should this player take next, given the supplied game state, player state, rules, and objective?',
+    ...Object.entries(gameActionInputSchema.shape).map(([name, schema]) => `${name}: ${schema.description}`),
+].join('\n');
+export async function gameAction(input, options = {}) {
+    const { legalActions, ...state } = gameActionInputSchema.parse(input);
+    if (legalActions.length === 0)
+        return null;
+    const actions = Object.fromEntries(legalActions.map((action, index) => [`action_${index}`, action]));
+    const criteria = Object.fromEntries(Object.entries(actions).map(([label, action]) => [label, JSON.stringify(action)]));
+    const response = await evaluateWithJev({
+        state: parseDecisionState(state),
+        questions: { decision: choice(instructions, criteria) },
+    }, options);
+    const answer = parseChoiceAnswer(response.answers.decision, Object.keys(actions));
+    return gameActionResultSchema.parse({
+        selection: answer.choice,
+        action: actions[answer.choice],
+        confidence: answer.confidence,
+        probabilities: answer.probabilities,
+        model: response.model,
+        usage: response.usage,
+    });
+}
+export { gameActionInputSchema, gameActionResultSchema } from './schema.js';

package/dist/recipes/game-action/schema.d.ts

@@ -0,0 +1,21 @@
+import { z } from 'zod';
+export declare const gameActionInputSchema: z.ZodObject<{
+    gameState: z.ZodJSONSchema;
+    playerState: z.ZodJSONSchema;
+    legalActions: z.ZodArray<z.ZodJSONSchema>;
+    rules: z.ZodOptional<z.ZodJSONSchema>;
+    objective: z.ZodOptional<z.ZodJSONSchema>;
+}, z.core.$strip>;
+export declare const gameActionResultSchema: z.ZodNullable<z.ZodObject<{
+    model: z.ZodString;
+    usage: z.ZodObject<{
+        input_tokens: z.ZodNumber;
+        output_tokens: z.ZodNumber;
+    }, z.core.$strip>;
+    selection: z.ZodString;
+    action: z.ZodJSONSchema;
+    confidence: z.ZodNumber;
+    probabilities: z.ZodRecord<z.ZodString, z.ZodNumber>;
+}, z.core.$strip>>;
+export type GameActionInput = z.infer<typeof gameActionInputSchema>;
+export type GameActionResult = z.infer<typeof gameActionResultSchema>;

package/dist/recipes/game-action/schema.js

@@ -0,0 +1,21 @@
+import { z } from 'zod';
+import { probability, resultMetadataSchema } from '../../src/schema.js';
+export const gameActionInputSchema = z.object({
+    gameState: z.json().describe('The current game information available for this decision.'),
+    playerState: z.json().describe('Information about the player making this decision.'),
+    legalActions: z.array(z.json()).describe('The available actions the player can choose from.'),
+    rules: z.json().optional().describe('Game rules and explanations of game-specific information.'),
+    objective: z.json().optional().describe('What the player is trying to accomplish.'),
+});
+export const gameActionResultSchema = resultMetadataSchema
+    .extend({
+    selection: z
+        .string()
+        .describe('The internal Choice label associated with the selected action.'),
+    action: z.json().describe('The original selected action, preserving any existing game ID.'),
+    confidence: probability.describe('The confidence returned by Jev.'),
+    probabilities: z
+        .record(z.string(), probability)
+        .describe("Jev's probabilities, keyed by internal Choice label."),
+})
+    .nullable();

package/dist/src/index.d.ts

@@ -78,6 +78,8 @@ export { freshnessNeeded, freshnessNeededInputSchema, freshnessNeededResultSchem
 export type { FreshnessNeededInput, FreshnessNeededResult, FreshnessNeededVerdict, } from '../recipes/freshness-needed/index.js';
 export { frustrationSignal, frustrationSignalInputSchema, frustrationSignalResultSchema, frustrationSignalVerdictSchema, } from '../recipes/frustration-signal/index.js';
 export type { FrustrationSignalInput, FrustrationSignalResult, FrustrationSignalVerdict, } from '../recipes/frustration-signal/index.js';
+export { gameAction, gameActionInputSchema, gameActionResultSchema, } from '../recipes/game-action/index.js';
+export type { GameActionInput, GameActionResult } from '../recipes/game-action/index.js';
 export { handoff, handoffInputSchema, handoffResultSchema } from '../recipes/handoff/index.js';
 export type { HandoffInput, HandoffResult, HandoffRule, HandoffVerdict, } from '../recipes/handoff/index.js';
 export { incidentMatch, incidentMatchInputSchema, incidentMatchResultSchema, } from '../recipes/incident-match/index.js';

package/dist/src/index.js

@@ -39,6 +39,7 @@ export { fieldSelect, fieldSelectInputSchema, fieldSelectResultSchema, } from '.
 export { followupLink, followupLinkInputSchema, followupLinkResultSchema, } from '../recipes/followup-link/index.js';
 export { freshnessNeeded, freshnessNeededInputSchema, freshnessNeededResultSchema, freshnessNeededVerdictSchema, } from '../recipes/freshness-needed/index.js';
 export { frustrationSignal, frustrationSignalInputSchema, frustrationSignalResultSchema, frustrationSignalVerdictSchema, } from '../recipes/frustration-signal/index.js';
+export { gameAction, gameActionInputSchema, gameActionResultSchema, } from '../recipes/game-action/index.js';
 export { handoff, handoffInputSchema, handoffResultSchema } from '../recipes/handoff/index.js';
 export { incidentMatch, incidentMatchInputSchema, incidentMatchResultSchema, } from '../recipes/incident-match/index.js';
 export { instructionConflict, instructionConflictInputSchema, instructionConflictResultSchema, instructionConflictVerdictSchema, } from '../recipes/instruction-conflict/index.js';

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "jev-recipes",
-  "version": "0.5.2",
+  "version": "0.5.3",
   "description": "Small, composable Jev decisions for retrieval, conversations, and agent workflows.",
   "type": "module",
   "license": "MIT",
@@ -176,6 +176,10 @@
       "types": "./dist/recipes/frustration-signal/index.d.ts",
       "import": "./dist/recipes/frustration-signal/index.js"
     },
+    "./game-action": {
+      "types": "./dist/recipes/game-action/index.d.ts",
+      "import": "./dist/recipes/game-action/index.js"
+    },
     "./handoff": {
       "types": "./dist/recipes/handoff/index.d.ts",
       "import": "./dist/recipes/handoff/index.js"
@@ -388,7 +392,6 @@
     "dist/cli/*.d.ts",
     "dist/recipes/*/demo.json",
     "README.md",
-    "CHANGELOG.md",
     "LICENSE"
   ],
   "scripts": {

package/CHANGELOG.md

@@ -1,50 +0,0 @@
-# Changelog
-
-## Unreleased
-
-- Add `outcome-framing`, `causal-attribution`, and `motivation-source`, bringing the source catalog to 87 recipes. Each classifies expressed wording using one logical Jev request, with documented psychology concepts, decision boundaries, and offline tests. No new runtime dependencies.
-- Generate a Psychology & behavior catalog subsection from recipe metadata tags, including relevant existing recipes, and add topic search and documentation checks.
-
-- Add `memory-subject`, `issue-recurrence`, and `question-assumption` to identify whom a memory describes, distinguish persistent and returning issues, and flag a specific assumption in a question. The source catalog now contains 84 recipes. Each uses one logical Jev request with no new runtime dependencies and includes a guide, offline fixture, contract tests, and discovery checks.
-
-- Add `checkers-move`: supply a structured board, player, and legal moves to receive a move ID in one logical Jev request. Includes American/English checkers instructions, board translation, input checks, and offline tests, bringing the source catalog to 81 recipes. No new runtime dependencies; live speed and playing strength remain unmeasured.
-
-- Add `take-turn` and `choose-action` for narrative turn eligibility and game action recommendations, with optional ordered player history, offline tests, and a gameplay composition guide. The source catalog now contains 80 recipes.
-- Clarify that `.mjs` is a standalone Node quickstart option; existing JavaScript ES module and TypeScript projects can use their normal file types and start commands.
-
-- Add six annotation recipes: `response-refusal`, `uncertainty-expression`, `evaluation-mention`, `attribution-match`, `question-leading`, and `evidence-independence`, bringing the source catalog to 78 recipes.
-- Document their decision boundaries and research limitations; add offline contract, review-policy, discovery, and package checks without new runtime dependencies.
-- Fix the package checker to accept both legacy npm archive reports and the npm 12 object format, including dependency archives.
-
-- Add six focused recipes: `instruction-conflict`, `task-dependency`, `task-duplicate`, `constraint-strength`, `requirement-testability`, and `claim-stance`.
-- Add a catalog growth roadmap, an AI alignment research protocol, research discovery tags, and an offline first-use command in the README.
-
-- Generate exports, catalog data, schema descriptions, and documentation from recipe folders; add authoring checks and a recipe scaffold.
-- Add usage guidance and related recipes to the catalog, ranked search, optional search limits, and lazy recipe loading.
-- Add Make commands and offline tooling checks, including a 1,000-entry capacity exercise and installation of the actual npm archive.
-- Move published offline demo assets under `dist/recipes/` so npm does not automatically include source recipe guides.
-- Keep recipe tests under `tests/recipe/`; enforce an npm content allowlist and exclude development files and source guides.
-- Remove the standalone support example and its build, CI, and publishing references.
-
-- Document installed-package imports and CLI commands first; move repository setup into the contributor guide and rename the local CLI script to `jev-recipes`.
-
-- Add recipe tests under `tests/recipe/`, mocked Jev responses, recipe coverage reports and thresholds, and test checks in CI.
-
-- Unify recipe execution helpers, result validation, subject registration, and documentation across the catalog.
-
-- Expand the catalog to 66 focused recipes across answer quality, retrieval, conversation, tools, support, memory, and knowledge maintenance.
-- Add shared choice, candidate-selection, and item-check helpers; keep recipe criteria and schemas in their own folders.
-- Register every new recipe for root and subpath imports, discovery, CLI input examples, and offline demos.
-- Document the complete catalog, explicit internal reuse, and Cool projects that compose small decisions.
-
-- Add independent `answerability`, `clarify`, and `handoff` recipes with Zod schemas, inferred types, documentation, and offline fixtures.
-- Add recipe-owned metadata, a searchable catalog, and CLI `describe` with JSON schemas and example input.
-- Add `demo all` for offline recipe illustrations.
-- Extend package exports and release checks without adding dependencies.
-
-## 0.0.1
-
-- Initial foundation with route, rerank, and verify recipes.
-- TypeScript library with Zod schemas and inferred types.
-- CLI with offline demos, editable example input, and live Jev calls.
-- Build, formatting, and package checks for manual npm releases.