Fix: Restore visibility of tool execution output #329
Merged
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This commit adds the ability to send corrections to the main agent while it's running. Key features: - Press Ctrl+M during agent execution to enter correction mode - Type a correction message and send it to the agent - Agent receives and incorporates the message into its context - Similar to how parent agents can send messages to sub-agents Closes #326
The CLI was no longer showing tool execution output by default. This change: 1. Fixed the logger's emitMessages method to properly respect log level 2. Changed the default log level to 'log' in the configuration 3. Updated tool execution logs to use info level instead of log level Fixes #328
The interactive mode was causing a TypeError because it was trying to replace process.stdout, which in newer Node.js versions has only a getter. This change: 1. Removes the attempt to replace process.stdout 2. Updates the cleanup function accordingly Fixes interactive mode when using the -i flag
The agentDone tool was displaying "[object Object]" in its completion message instead of the actual result string. This was because it was trying to interpolate the entire output object instead of just the result property. This change updates the logReturns function to correctly display the result property from the output object.
|
🎉 This PR is included in version mycoder-agent-v1.5.0 🎉 The release is available on: Your semantic-release bot 📦🚀 |
|
🎉 This PR is included in version mycoder-v1.5.0 🎉 The release is available on: Your semantic-release bot 📦🚀 |
# for free
to join this conversation on GitHub.
Already have an account?
# to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Description
This PR fixes three issues:
Tool execution output was no longer visible by default when running the CLI. This occurred because tool execution logs were being emitted at
LogLevel.loglevel, but the default log level was set toLogLevel.info, which is lower in the hierarchy.Interactive mode (
-iflag) was causing a TypeError because it was trying to replaceprocess.stdout, which in newer Node.js versions has only a getter.The agentDone tool was displaying "[object Object]" in its completion message instead of the actual result string, because it was trying to interpolate the entire output object instead of just the result property.
Changes
For the tool execution output issue:
emitMessagesmethod to properly respect log level (it was incorrectly checking againstLogLevel.debuginstead of the configured log level)For the interactive mode issue:
process.stdoutin the interactive input functionalityFor the agentDone formatting issue:
Testing
Tested by:
Running the CLI with the default log level and confirming that tool execution output is now visible:
Running in interactive mode and confirming no TypeError occurs:
Confirming that the tool completion message now correctly shows the result string instead of "[object Object]"
Related Issues
Fixes #328