Issue Station
Sign In
Tools hereToolbox

claude-code - 💡(How to fix) Fix VS Code Extension: Real-time activity panel for tool call visibility [1 participants]

Official PRs (…)
ON THIS PAGE

Recommended Tools

×6

Utilities matched from this issue’s tags and category — try them while you read without losing context.

GitHub issue graph ai analysis

Paste a GitHub issue URL. We fetch that issue, discover linked issues from bodies/comments/timeline, collect linked pull requests, and produce a structured English report.

The report is written in English Markdown for sharing and archival.

Helpful · Quick feedback

Loading…
GitHub stats
anthropics/claude-code#49631•Fetched 2026-04-17 08:35:43
View on GitHub
Comments
0
Participants
1
Timeline
3
Reactions
0
Author
sridharankaliyamoorthysridharankaliyamoorthy
Participants
sridharankaliyamoorthysridharankaliyamoorthy
Timeline (top)
labeled ×3

The terminal CLI already shows this level of detail when I use the claude --verbose or claude --debug commands. The VS Code extension should work the way.

As someone who switched from Cursor after a year and a half this is the difference in user experience for me. Everything else, about the VS Code extension is better. It uses tokens efficiently has a better context window and the model is better.. This one feature would make it complete.

Root Cause

The terminal CLI already shows this level of detail when I use the claude --verbose or claude --debug commands. The VS Code extension should work the way.

As someone who switched from Cursor after a year and a half this is the difference in user experience for me. Everything else, about the VS Code extension is better. It uses tokens efficiently has a better context window and the model is better.. This one feature would make it complete.

RAW_BUFFERClick to expand / collapse

Problem

When I use Claude Code in the VS Code extension I do not have a view of what the AI agent is doing. The agent makes tool calls, reads and edits files and runs commands but all of this happens behind the scenes. I only see the result in the chat panel.

In comparison tools like Cursor show what is happening in the terminal in time. This makes it clear what the AI is doing. Gives me confidence in the process.

This is especially important for three things:

  • Trust. I want to see what the agent is reading, writing and executing in time

  • Debugging. I need to understand why something did not work

  • Learning. I want to watch how the agent solves problems

Proposed Solution

I think the VS Code extension should have a real-time activity panel*. This panel would show what is happening as it happens like the CLIs verbose or debug mode. This panel should show:

  1. Tool calls as they happen. File reads, edits and bash commands

  2. File paths being accessed. Which files the agent is reading or writing

  3. Terminal commands. The exact commands being run and what they output

  4. ** Usage**. How many tokens are used for each message

  5. Timing. How long each operation takes

UX Suggestion

  • I think there should be a Activity" tab in the Claude Code sidebar next to the chat panel

  • Or maybe a panel at the bottom that can be turned on or off like a terminal

  • This should be something users can choose to turn on so the interface stays clean for users who do not need it

Context

The terminal CLI already shows this level of detail when I use the claude --verbose or claude --debug commands. The VS Code extension should work the way.

As someone who switched from Cursor after a year and a half this is the difference in user experience for me. Everything else, about the VS Code extension is better. It uses tokens efficiently has a better context window and the model is better.. This one feature would make it complete.

Environment

  • I am using the Claude Code VS Code Extension

  • My operating system is macOS, which is based on Darwin

  • I have the Max plan, which's 20 times the normal plan

extent analysis

TL;DR

The proposed solution involves adding a real-time activity panel to the VS Code extension to provide transparency into the AI agent's actions, similar to the verbose or debug mode in the CLI.

Guidance

  • To achieve this, the extension could utilize the existing verbose or debug logging functionality from the CLI and display it in a dedicated panel within the VS Code interface.
  • The panel should be optional and toggleable to maintain a clean interface for users who do not require this level of detail.
  • The panel's content should include tool calls, file paths being accessed, terminal commands with their outputs, token usage, and operation timings to address the user's needs for trust, debugging, and learning.
  • The extension's developers could consider adding a setting to control the level of detail displayed in the activity panel, allowing users to customize the output based on their preferences.

Example

No specific code example can be provided without more context on the extension's internal implementation. However, the idea would be to capture and display the output from the claude --verbose or claude --debug commands in a panel within the VS Code extension.

Notes

The implementation details would depend on how the VS Code extension is currently structured and how it interacts with the Claude AI agent. This solution assumes that the necessary logging or debugging information is already available within the extension or can be easily accessed.

Recommendation

Apply a workaround by running the CLI with verbose or debug mode in a separate terminal to observe the agent's actions until the real-time activity panel is implemented in the VS Code extension. This provides an immediate, albeit less integrated, solution to the user's need for transparency into the AI agent's activities.

Vote matrix · Quick signals

Works
Did the solution work? Tap to confirm.
Easy Fix
Was it a quick fix?
Time Saver
Did it save you time?
Blocking
Was it severely blocking?
Common Issue
Are others likely hitting this too?
Flaky / Intermittent
Is it intermittent?
Verified / Reproducible
Can you reproduce it reliably?
Loading…
#generation error#database connection#vector store#embedding generation#cache error

Still need to ship something?

×6

Another batch ranked right after the header list — different links, same matching logic.

Back to top recommendations

TRENDING