@statelyai/agent 2.0.0-next.1 → 2.0.0-next.3

package/.changeset/grumpy-dolphins-think.md

@@ -0,0 +1,17 @@
+---
+'@statelyai/agent': minor
+---
+
+planner -> strategy
+agent.addPlan -> agent.addDecision
+agent.getPlans -> agent.getDecisions
+
+The word "strategy" is now used instead of "planner" to make it more clear what the agent is doing: it uses a strategy to make decisions. The method `agent.addPlan(…)` has been renamed to `agent.addDecision(…)` and `agent.getPlans(…)` has been renamed to `agent.getDecisions(…)` to reflect this change. Additionally, you specify the `strategy` instead of the `planner` when creating an agent:
+
+```diff
+const agent = createAgent({
+- planner: createSimplePlanner(),
++ strategy: createSimpleStrategy(),
+  ...
+});
+```

package/.changeset/old-teachers-tap.md

@@ -0,0 +1,5 @@
+---
+'@statelyai/agent': minor
+---
+
+For feedback, the `goal`, `observationId`, and `attributes` are now required, and `feedback` and `reward` are removed since they are redundant.

package/.changeset/pink-eagles-deliver.md

@@ -0,0 +1,13 @@
+---
+'@statelyai/agent': patch
+---
+
+The `score` is now required for feedback:
+
+```ts
+agent.addFeedback({
+  score: 0.5,
+  goal: 'Win the game',
+  observationId: '...',
+});
+```

package/.changeset/pre.json

@@ -7,7 +7,15 @@
   "changesets": [
     "cyan-carpets-perform",
     "fast-donkeys-argue",
+    "grumpy-dolphins-think",
     "light-hats-drive",
-    "old-jobs-check"
+    "old-jobs-check",
+    "old-teachers-tap",
+    "pink-eagles-deliver",
+    "quiet-turtles-do",
+    "smart-yaks-pull",
+    "sweet-clouds-mix",
+    "swift-mangos-rush",
+    "tough-ways-rhyme"
   ]
 }

package/.changeset/quiet-turtles-do.md

@@ -0,0 +1,7 @@
+---
+'@statelyai/agent': major
+---
+
+The `state` can no longer be specified in `agent.interact(...)`, since the actual state value is already observed and passed to the `strategy` function.
+
+The `context` provided to agent decision functions, like `agent.decide({ context })` and in `agent.interact(...)`, is now used solely to override the `state.context` provided to the prompt template.

package/.changeset/smart-yaks-pull.md

@@ -0,0 +1,23 @@
+---
+'@statelyai/agent': minor
+---
+
+You can specify `allowedEvents` in `agent.decide(...)` to allow from a list of specific events to be sent to the agent. This is useful when using `agent.decide(...)` without a state machine.
+
+```ts
+const agent = createAgent({
+  // ...
+  events: {
+    PLAY: z.object({}).describe('Play a move'),
+    SKIP: z.object({}).describe('Skip a move'),
+    FORFEIT: z.object({}).describe('Forfeit the game'),
+  },
+});
+
+// ...
+const decision = await agent.decide({
+  // Don't allow the agent to send `FORFEIT` or other events
+  allowedEvents: ['PLAY', 'SKIP'],
+  // ...
+});
+```

package/.changeset/sweet-clouds-mix.md

@@ -0,0 +1,16 @@
+---
+'@statelyai/agent': patch
+---
+
+The entire observed `state` must be provided, instead of only `context`, for any agent decision making functions:
+
+```ts
+agent.interact(actor, (obs) => {
+  // ...
+  return {
+    goal: 'Some goal',
+    // instead of context
+    state: obs.state,
+  };
+});
+```

package/.changeset/swift-mangos-rush.md

@@ -0,0 +1,5 @@
+---
+"@statelyai/agent": patch
+---
+
+Remove `goal` from feedback input

package/.changeset/tough-ways-rhyme.md

@@ -0,0 +1,5 @@
+---
+"@statelyai/agent": minor
+---
+
+Add `score` and `comment` fields for feedback

package/CHANGELOG.md

@@ -1,5 +1,84 @@
 # @statelyai/agent
 
+## 2.0.0-next.3
+
+### Major Changes
+
+- [`bf6b468`](https://github.com/statelyai/agent/commit/bf6b468d66d58bf53629d70d1b2a273948c9ba1e) Thanks [@davidkpiano](https://github.com/davidkpiano)! - The `state` can no longer be specified in `agent.interact(...)`, since the actual state value is already observed and passed to the `strategy` function.
+
+  The `context` provided to agent decision functions, like `agent.decide({ context })` and in `agent.interact(...)`, is now used solely to override the `state.context` provided to the prompt template.
+
+### Minor Changes
+
+- [`1287a6d`](https://github.com/statelyai/agent/commit/1287a6d405ed3bd6be37a61aaa9d54d963b5b1cd) Thanks [@davidkpiano](https://github.com/davidkpiano)! - Add `score` and `comment` fields for feedback
+
+### Patch Changes
+
+- [`5b5a8e7`](https://github.com/statelyai/agent/commit/5b5a8e7012d550f5b05d0fefc9eade7731202577) Thanks [@davidkpiano](https://github.com/davidkpiano)! - The `score` is now required for feedback:
+
+  ```ts
+  agent.addFeedback({
+    score: 0.5,
+    goal: "Win the game",
+    observationId: "...",
+  });
+  ```
+
+- [`5b5a8e7`](https://github.com/statelyai/agent/commit/5b5a8e7012d550f5b05d0fefc9eade7731202577) Thanks [@davidkpiano](https://github.com/davidkpiano)! - The entire observed `state` must be provided, instead of only `context`, for any agent decision making functions:
+
+  ```ts
+  agent.interact(actor, (obs) => {
+    // ...
+    return {
+      goal: "Some goal",
+      // instead of context
+      state: obs.state,
+    };
+  });
+  ```
+
+- [`9d65d71`](https://github.com/statelyai/agent/commit/9d65d71e41f9f0f84f637ea7e0a0e22ccf67f264) Thanks [@davidkpiano](https://github.com/davidkpiano)! - Remove `goal` from feedback input
+
+## 2.0.0-next.2
+
+### Minor Changes
+
+- [`4d870fe`](https://github.com/statelyai/agent/commit/4d870fe38ad0c906bafb2e0f6b2dabb745900ad3) Thanks [@davidkpiano](https://github.com/davidkpiano)! - planner -> strategy
+  agent.addPlan -> agent.addDecision
+  agent.getPlans -> agent.getDecisions
+
+  The word "strategy" is now used instead of "planner" to make it more clear what the agent is doing: it uses a strategy to make decisions. The method `agent.addPlan(…)` has been renamed to `agent.addDecision(…)` and `agent.getPlans(…)` has been renamed to `agent.getDecisions(…)` to reflect this change. Additionally, you specify the `strategy` instead of the `planner` when creating an agent:
+
+  ```diff
+  const agent = createAgent({
+  - planner: createSimplePlanner(),
+  + strategy: createSimpleStrategy(),
+    ...
+  });
+  ```
+
+- [`f1189cb`](https://github.com/statelyai/agent/commit/f1189cb980e52fa909888d27d3300dcd913ea47f) Thanks [@davidkpiano](https://github.com/davidkpiano)! - For feedback, the `goal`, `observationId`, and `attributes` are now required, and `feedback` and `reward` are removed since they are redundant.
+
+- [`7b16326`](https://github.com/statelyai/agent/commit/7b163266c61bfc8125ed4d00924680d932001e27) Thanks [@davidkpiano](https://github.com/davidkpiano)! - You can specify `allowedEvents` in `agent.decide(...)` to allow from a list of specific events to be sent to the agent. This is useful when using `agent.decide(...)` without a state machine.
+
+  ```ts
+  const agent = createAgent({
+    // ...
+    events: {
+      PLAY: z.object({}).describe("Play a move"),
+      SKIP: z.object({}).describe("Skip a move"),
+      FORFEIT: z.object({}).describe("Forfeit the game"),
+    },
+  });
+
+  // ...
+  const decision = await agent.decide({
+    // Don't allow the agent to send `FORFEIT` or other events
+    allowedEvents: ["PLAY", "SKIP"],
+    // ...
+  });
+  ```
+
 ## 2.0.0-next.1
 
 ### Minor Changes