From da28681f042e2f0fcf22e279da28b28a03eebf72 Mon Sep 17 00:00:00 2001 From: Cursor Agent Date: Wed, 12 Aug 2026 00:54:39 +0000 Subject: [PATCH] docs: correct kickoff path to AgentExecutor agent.kickoff() uses AgentExecutor and returns LiteAgentOutput; it does not run through a LiteAgent class. Update Secure Agent Design and the Agents concept note accordingly. Co-authored-by: Rip&Tear --- docs/edge/en/concepts/agents.mdx | 7 ++++--- docs/edge/en/guides/agents/secure-agent-design.mdx | 2 +- 2 files changed, 5 insertions(+), 4 deletions(-) diff --git a/docs/edge/en/concepts/agents.mdx b/docs/edge/en/concepts/agents.mdx index 98fffbf6e..7f1f71df2 100644 --- a/docs/edge/en/concepts/agents.mdx +++ b/docs/edge/en/concepts/agents.mdx @@ -638,9 +638,10 @@ asyncio.run(main()) ``` - The `kickoff()` method uses a `LiteAgent` internally, which provides a simpler - execution flow while preserving all of the agent's configuration (role, goal, - backstory, tools, etc.). + The `kickoff()` method uses an `AgentExecutor` directly (no Task or Crew), + which provides a simpler execution flow while preserving all of the agent's + configuration (role, goal, backstory, tools, etc.). It returns a + `LiteAgentOutput`. ## Important Considerations and Best Practices diff --git a/docs/edge/en/guides/agents/secure-agent-design.mdx b/docs/edge/en/guides/agents/secure-agent-design.mdx index 7d574c818..b6eb038b0 100644 --- a/docs/edge/en/guides/agents/secure-agent-design.mdx +++ b/docs/edge/en/guides/agents/secure-agent-design.mdx @@ -24,7 +24,7 @@ Use this guide whenever an agent touches user data, external content, or side-ef ### Single-agent `kickoff()` -`agent.kickoff(...)` runs through a LiteAgent — no Task and no Crew. Controls differ: +`agent.kickoff(...)` runs through an `AgentExecutor` — no Task and no Crew. Controls differ: | Still applies | Does **not** apply | | --- | --- |