Restore optional MCP Context7 configuration hooks#168
Conversation
![chatgpt-codex-connector[bot]](https://avatars.githubusercontent.com/in/1144995?s=60&v=4){kind=small}
There was a problem hiding this comment.
💡 Codex Review
Here are some automated review suggestions for this pull request.
ℹ️ About Codex in GitHub
Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you
- Open a pull request for review
- Mark a draft as ready
- Comment "@codex review".
If Codex has suggestions, it will comment; otherwise it will react with 👍.
Codex can also answer questions or update the PR. Try commenting
| } | ||
|
|
||
| if let Some(tool) = tool_name { | ||
| if tool.starts_with("mcp_context7") { | ||
| render_mcp_context7_output(renderer, val)?; | ||
| } else if tool.starts_with("mcp_sequentialthinking") { | ||
| if tool.starts_with("mcp_sequentialthinking") { | ||
| render_mcp_sequential_output(renderer, val)?; | ||
| } | ||
| } |
There was a problem hiding this comment.
The previous version had a render_mcp_context7_output branch that displayed messages, errors, and metadata for Context7 tools. This branch was removed, and render_tool_output now only special-cases sequential-thinking tools while the remainder of the function only prints stdout/stderr. Context7 MCP replies do not include those fields, so enabling the optional provider from vtcode.toml will now produce no visible output for successful lookups, making the integration effectively unusable. Consider restoring a renderer or at least serializing the payload so users can see the results.
Useful? React with 👍 / 👎.
There was a problem hiding this comment.
💡 Codex Review
Here are some automated review suggestions for this pull request.
ℹ️ About Codex in GitHub
Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you
- Open a pull request for review
- Mark a draft as ready
- Comment "@codex review".
If Codex has suggestions, it will comment; otherwise it will react with 👍.
Codex can also answer questions or update the PR. Try commenting
| if let Some(tool) = tool_name { | ||
| if tool.starts_with("mcp_context7") { | ||
| render_mcp_context7_output(renderer, val)?; | ||
| } else if tool.starts_with("mcp_sequentialthinking") { | ||
| if tool.starts_with("mcp_sequentialthinking") { | ||
| render_mcp_sequential_output(renderer, val)?; | ||
| } |
There was a problem hiding this comment.
The special-case handling for mcp_context7* tools was removed, so render_tool_output now only checks for mcp_sequentialthinking*. Context7 responses don’t populate stdout/stderr; they emit a JSON body with fields like status, messages, and errors. With the branch gone, the renderer prints nothing when the optional Context7 provider is enabled, which makes the tool appear to return no data. Reintroducing a renderer (or a generic JSON fallback) is necessary so Context7 results remain visible.
Useful? React with 👍 / 👎.
|
@codex fix comments |
|
Summary
Testing
|
1 participant
Summary
Testing
https://chatgpt.com/codex/tasks/task_e_68e4e23002288323961c525ccb257bc9