fix: require AgentSwarm to run alone

[kermanx]

Related Issue

No linked issue. This PR addresses the need to keep AgentSwarm execution exclusive within a model response.

Problem

AgentSwarm launches many subagents from one tool call. If a model emits AgentSwarm alongside other tools, or emits multiple AgentSwarm calls in the same response, the current batch can start overlapping side effects before the model has observed the first swarm result.

What changed

• Added a permission policy that denies any tool-call batch containing AgentSwarm unless the batch contains exactly one AgentSwarm call.
• Passed the full same-response toolCalls list through tool execution hook contexts so permission policies can make batch-level decisions.
• Updated the AgentSwarm tool description and bilingual tool reference docs to state that an AgentSwarm call must be the only tool call in the response.
• Added patch changesets for @moonshot-ai/agent-core and @moonshot-ai/kimi-code.

Checklist

• I have read the CONTRIBUTING document.
• I have linked a related issue, or explained the problem above.
• I have added tests that prove my feature works.
• Ran gen-changesets skill, or this PR needs no changeset.
• Ran gen-docs skill, or this PR needs no doc update.

Verification:

• pnpm --filter @moonshot-ai/agent-core typecheck
• pnpm --filter @moonshot-ai/agent-core test
• pnpm --filter kimi-code-docs build

[changeset-bot]

🦋 Changeset detected

Latest commit: 5dd51ca

The changes in this PR will be included in the next version bump.

This PR includes changesets to release 2 packages

Name	Type
@moonshot-ai/agent-core	Patch
@moonshot-ai/kimi-code	Patch

Not sure what this means? Click here to learn what changesets are.

Click here if you're a maintainer who wants to add another changeset to this PR

[pkg-pr-new]

pnpm dlx https://pkg.pr.new/@moonshot-ai/kimi-code@5dd51ca

npx https://pkg.pr.new/@moonshot-ai/kimi-code@5dd51ca

commit: 5dd51ca