Skip to content

Commit e36b771

Browse files
authored
docs: document function tool execution concurrency (#1259)
1 parent 4501278 commit e36b771

1 file changed

Lines changed: 4 additions & 0 deletions

File tree

docs/src/content/docs/guides/running-agents.mdx

Lines changed: 4 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -70,6 +70,7 @@ The additional options are:
7070
| `reasoningItemIdPolicy` || Controls whether reasoning-item `id`s are preserved or omitted when prior run items are turned back into model input. See [Reasoning item ID policy](#reasoning-item-id-policy). |
7171
| `tracing` || Per-run tracing configuration overrides (for example, export API key). |
7272
| `sandbox` || Sandbox client, live session, session state, snapshot, manifest override, or concurrency limits for `SandboxAgent` runs. See [Sandbox agents](/openai-agents-js/guides/sandbox-agents/concepts). |
73+
| `toolExecution` || SDK-side execution settings for local tool calls. Use `toolExecution.maxFunctionToolConcurrency` to limit how many function tools run at once. |
7374
| `errorHandlers` || Handlers for supported runtime errors (currently `maxTurns`). See [Error handlers](#error-handlers). |
7475
| `conversationId` || Reuse a server-side conversation (OpenAI Responses API + Conversations API only). |
7576
| `previousResponseId` || Continue from the previous Responses API call without creating a conversation (OpenAI Responses API only). |
@@ -101,6 +102,9 @@ If you are creating your own `Runner` instance, you can pass in a `RunConfig` ob
101102
| `toolErrorFormatter` | `ToolErrorFormatter` | Global hook to customize tool approval rejection messages returned to the model. |
102103
| `reasoningItemIdPolicy` | `ReasoningItemIdPolicy` | Default policy for preserving or omitting reasoning-item `id`s when replaying generated items into later model calls. |
103104
| `sandbox` | `SandboxRunConfig` | Default sandbox runtime configuration for `SandboxAgent` runs. |
105+
| `toolExecution` | `ToolExecutionConfig` | Default SDK-side execution settings for local tool calls. `maxFunctionToolConcurrency` caps local function tool concurrency for each turn; unset or `null` starts all function tool calls emitted in a turn. |
106+
107+
`toolExecution.maxFunctionToolConcurrency` must be an integer greater than or equal to `1`. This setting only limits SDK-side execution of local function tools. It does not change provider-side `modelSettings.parallelToolCalls`.
104108

105109
## State and conversation management
106110

0 commit comments

Comments
 (0)