Issue #, if available:
N/A

Description of changes:
• Adds configurable tool output verbosity via new setting chat.tool_output_verbose (default: true)
• When set to false, hides detailed tool execution output for trusted tools while preserving:
• Permission prompts for untrusted tools (safety preserved)
• Completion messages with tool names: ● Completed in 0.99s (fs_read, fs_write)
• Structured events for IDE integration
• Implementation changes:
• Added ChatToolOutputVerbose setting to Setting enum with full mapping
• Added tools_used_in_turn: Vec<String> field to ChatSession to track tools used in current turn
• Modified print_tool_description() to conditionally display output based on verbose || !trusted
• Enhanced completion message to append tool names when not verbose
• Added show_details parameter to Tool::invoke() and propagated through to queue_function_result()
• Updated fs_read tool implementations to respect show_details flag (6 call sites)

Tests

Test case: Default behavior (verbose=true)

List files in current directory

Output shows full tool details:

🛠️  Using tool: fs_read (trusted)
 ⋮ 
 ● Reading directory: /path/to/dir with maximum depth of 1
 ✓ Successfully read directory /path/to/dir (32 entries)
 ⋮ 
 ● Completed in 0.99s

Test case: Reduced verbosity (verbose=false, trusted tools)

List files in current directory

Output shows only completion: ● Completed in 0.99s (fs_read)

Test case: Permission prompts preserved (verbose=false, untrusted tools)

Create a test directory

Output shows full details with permission prompt:

🛠️  Using tool: execute_bash
 ⋮ 
 ● I will run the following shell command: mkdir -p /tmp/test
 ⋮ 
Allow this action? [y/n/t]:

Test case: Multiple tools (verbose=false)

Read README.md and create a summary

Output shows all tools used: ● Completed in 1.5s (fs_read, fs_write)

By submitting this pull request, I confirm that you can use, modify, copy, and redistribute this contribution, under the terms of your choice.