Merge branch 'main' into feat/mcp-tool-filtering-js

Author: seratch

.changeset/funny-dragons-tell.md

@@ -0,0 +1,5 @@
+---
+'@openai/agents-core': patch
+---
+
+Rebinds cached tools to the current MCP server to avoid stale tool invocation (fixes #195)

.changeset/hungry-ladybugs-lick.md

@@ -0,0 +1,5 @@
+---
+"@openai/agents-openai": patch
+---
+
+fix: if prompt is not specified return undefined - fixes #159

.changeset/shiny-beers-tie.md

@@ -0,0 +1,5 @@
+---
+'@openai/agents-openai': patch
+---
+
+Add input_fidelity parameter support to image generation tool

.changeset/strong-lobsters-repair.md

@@ -0,0 +1,6 @@
+---
+"@openai/agents-extensions": patch
+"@openai/agents-realtime": patch
+---
+
+fix(realtime-ws): stop accidental cancellation error

.changeset/tame-deers-shake.md

@@ -0,0 +1,7 @@
+---
+"@openai/agents-core": patch
+"@openai/agents-openai": patch
+"@openai/agents-realtime": patch
+---
+
+Fix typos across repo

.changeset/young-birds-hammer.md

@@ -0,0 +1,6 @@
+---
+'@openai/agents-openai': patch
+'@openai/agents-core': patch
+---
+
+Fix #216 Publicly accessible PDF file URL is not yet supported in the input_file content data

.github/workflows/test.yml

@@ -13,7 +13,8 @@ jobs:
     strategy:
       matrix:
         # https://nodejs.org/en/about/previous-releases
-        node-version: [20, 22, 24]
+        # Using 24.3.x due to https://github.com/pnpm/pnpm/issues/9743
+        node-version: ['20', '22', '24.3.x']
     steps:
       - name: Checkout repository
         uses: actions/checkout@v4
@@ -36,4 +37,4 @@ jobs:
       - name: Compile examples
         run: pnpm -r build-check
       - name: Run tests
-        run: pnpm test
+        run: pnpm test

docs/src/content/docs/extensions/twilio.mdx

@@ -12,7 +12,7 @@ raw audio from a phone call to a WebSocket server. This set up can be used to co
 [voice agents](/openai-agents-js/guides/voice-agents) to Twilio. You can use the default Realtime Session transport
 in `websocket` mode to connect the events coming from Twilio to your Realtime Session. However,
 this requires you to set the right audio format and adjust your own interruption timing as phone
-calls will naturally introduce more latency than a web-based converstaion.
+calls will naturally introduce more latency than a web-based conversation.
 
 To improve the set up experience, we've created a dedicated transport layer that handles the
 connection to Twilio for you, including handling interruptions and audio forwarding for you.
@@ -72,7 +72,7 @@ for more information on how to use the `RealtimeSession` with voice agents.
 
    In order to receive all the necessary events and audio from Twilio, you should create your
    `TwilioRealtimeTransportLayer` instance as soon as you have a reference to the WebSocket
-   connetion and immediately call `session.connect()` afterwards.
+   connection and immediately call `session.connect()` afterwards.
 
 2. **Access the raw Twilio events.**
 

docs/src/content/docs/guides/mcp.mdx

@@ -59,7 +59,7 @@ To stream incremental MCP results, pass `stream: true` when you run the `Agent`:
 
 For sensitive operations you can require human approval of individual tool calls. Pass either `requireApproval: 'always'` or a fine‑grained object mapping tool names to `'never'`/`'always'`.
 
-If you can programatically determine whether a tool call is safe, you can use the [`onApproval` callback](https://github.com/openai/openai-agents-js/blob/main/examples/mcp/hosted-mcp-on-approval.ts) to approve or reject the tool call. If you require human approval, you can use the same [human-in-the-loop (HITL) approach](/openai-agents-js/guides/human-in-the-loop/) using `interruptions` as for local function tools.
+If you can programmatically determine whether a tool call is safe, you can use the [`onApproval` callback](https://github.com/openai/openai-agents-js/blob/main/examples/mcp/hosted-mcp-on-approval.ts) to approve or reject the tool call. If you require human approval, you can use the same [human-in-the-loop (HITL) approach](/openai-agents-js/guides/human-in-the-loop/) using `interruptions` as for local function tools.
 
 <Code
   lang="typescript"

docs/src/content/docs/guides/results.mdx

@@ -46,7 +46,7 @@ There are two ways you can access the inputs for your next turn:
 
 ## Last agent
 
-The `lastAgent` property contains the last agent that ran. Depending on your application, this is often useful for the next time the user inputs something. For example, if you have a frontline triage agent that hands off to a language-specific agent, you can store the last agent, and re-use it the next time the user messages the agent.
+The `lastAgent` property contains the last agent that ran. Depending on your application, this is often useful for the next time the user inputs something. For example, if you have a frontline triage agent that hands off to a language-specific agent, you can store the last agent, and reuse it the next time the user messages the agent.
 
 In streaming mode it can also be useful to access the `currentAgent` property that's mapping to the current agent that is running.