Conversation
|
There was a problem hiding this comment.
1 issue found and verified against the latest diff
Confidence score: 4/5
- In
packages/docs/v4/integrations/deep-agents.mdx, the integration introduction leads with framework details instead of the reader-facing capability, which may make the guide less clear and compelling; lead with the capability, then explain LangChain MCP adapters as the implementation.
Prompt for AI agents (unresolved issues)
Check if these issues are valid — if so, understand the root cause of each and fix them. If appropriate, use sub-agents to investigate and fix each issue separately.
<file name="packages/docs/v4/integrations/deep-agents.mdx">
<violation number="1" location="packages/docs/v4/integrations/deep-agents.mdx:19">
P2: Custom agent: **Stagehand docs prose guide**
Lead this integration introduction with the reader-facing capability, then identify LangChain MCP adapters as the implementation. The current opening leads with framework and adapter context before explaining what Stagehand enables.</violation>
</file>
Reply with feedback, questions, or to request a fix.
Re-trigger cubic
|
|
||
| ## Add Stagehand to an existing agent | ||
|
|
||
| Use Deep Agents’s supported tool integration: LangChain MCP adapters with an explicit persistent session. This is an [official SDK/MCP extension point](https://docs.langchain.com/oss/python/langchain/mcp); a separate host plugin marketplace is not required. |
There was a problem hiding this comment.
P2: Custom agent: Stagehand docs prose guide
Lead this integration introduction with the reader-facing capability, then identify LangChain MCP adapters as the implementation. The current opening leads with framework and adapter context before explaining what Stagehand enables.
Prompt for AI agents
Check if this issue is valid — if so, understand the root cause and fix it. At packages/docs/v4/integrations/deep-agents.mdx, line 19:
<comment>Lead this integration introduction with the reader-facing capability, then identify LangChain MCP adapters as the implementation. The current opening leads with framework and adapter context before explaining what Stagehand enables.</comment>
<file context>
@@ -11,9 +11,64 @@ The Deep Agents integration provides two deployment patterns with the same `run`
+## Add Stagehand to an existing agent
+
+Use Deep Agents’s supported tool integration: LangChain MCP adapters with an explicit persistent session. This is an [official SDK/MCP extension point](https://docs.langchain.com/oss/python/langchain/mcp); a separate host plugin marketplace is not required.
+
+<Note>
</file context>
| Use Deep Agents’s supported tool integration: LangChain MCP adapters with an explicit persistent session. This is an [official SDK/MCP extension point](https://docs.langchain.com/oss/python/langchain/mcp); a separate host plugin marketplace is not required. | |
| Use persistent Stagehand tools in your existing agent to run browser actions, capture snapshots, and take screenshots. LangChain MCP adapters provide this integration through an explicit session. |
There was a problem hiding this comment.
1 issue found and verified against the latest diff
Confidence score: 4/5
packages/docs/v4/integrations/deep-agents.mdxuses passive voice for the Node.js requirement, which may make the setup guidance less clear to readers; revise it to explicitly state that users need Node.js 24+ to run the MCP server.
Prompt for AI agents (unresolved issues)
Check if these issues are valid — if so, understand the root cause of each and fix them. If appropriate, use sub-agents to investigate and fix each issue separately.
<file name="packages/docs/v4/integrations/deep-agents.mdx">
<violation number="1" location="packages/docs/v4/integrations/deep-agents.mdx:22">
P2: Custom agent: **Stagehand docs prose guide**
This limitation uses passive voice without an explicit actor. Write `You need Node.js 24+ to run the MCP server.`</violation>
</file>
Reply with feedback, questions, or to request a fix.
Re-trigger cubic
| Use Deep Agents’s supported tool integration: LangChain MCP adapters with an explicit persistent session. This is an [official SDK/MCP extension point](https://docs.langchain.com/oss/python/langchain/mcp); you do not need a separate host marketplace plugin. | ||
|
|
||
| <Note> | ||
| The portable setup below requires the first `@browserbasehq/stagehand-mcp@0.1.0` release. Until that release is published, use the source example further down this page. Node.js 24+ is required to run the MCP server. |
There was a problem hiding this comment.
P2: Custom agent: Stagehand docs prose guide
This limitation uses passive voice without an explicit actor. Write You need Node.js 24+ to run the MCP server.
Prompt for AI agents
Check if this issue is valid — if so, understand the root cause and fix it. At packages/docs/v4/integrations/deep-agents.mdx, line 22:
<comment>This limitation uses passive voice without an explicit actor. Write `You need Node.js 24+ to run the MCP server.`</comment>
<file context>
@@ -11,9 +11,64 @@ The Deep Agents integration provides two deployment patterns with the same `run`
+Use Deep Agents’s supported tool integration: LangChain MCP adapters with an explicit persistent session. This is an [official SDK/MCP extension point](https://docs.langchain.com/oss/python/langchain/mcp); a separate host plugin marketplace is not required.
+
+<Note>
+The portable setup below requires the first `@browserbasehq/stagehand-mcp@0.1.0` release. Until that release is published, use the source example further down this page. Node.js 24+ is required to run the MCP server.
+</Note>
+
</file context>
| The portable setup below requires the first `@browserbasehq/stagehand-mcp@0.1.0` release. Until that release is published, use the source example further down this page. Node.js 24+ is required to run the MCP server. | |
| The portable setup below requires the first `@browserbasehq/stagehand-mcp@0.1.0` release. Until that release is published, use the source example further down this page. You need Node.js 24+ to run the MCP server. |
The Deep Agents integration already has a supported host tool API: LangChain MCP adapters. Lead the public guide with that integration route and a portable configuration for an existing application, instead of requiring a full Stagehand checkout.
This is an SDK/MCP integration, not a separate marketplace plugin. The source example remains available. The example holds an explicit persistent ClientSession so browser state and snapshot IDs survive across calls; the Python-native and Managed Deep Agents examples remain available.
Stacked on #2971 for the shared
@browserbasehq/stagehand-mcppackage. The first 0.1.0 release is required before the portable npm command is available; the guide explicitly identifies this prerequisite.E2E Test Matrix
Documentation validation: a combined snapshot of all nine updated integration guides passed
mint validateandmint broken-links --check-anchors --check-redirects --check-snippets.