Claude Code 2.1.149: Enhanced Usage Tracking and Better PowerShell Security for Indian Teams
Claude Code 2.1.149 introduces detailed usage breakdowns showing exactly what's consuming your API credits across skills, agents, and plugins, along with critical PowerShell security fixes. The update particularly benefits Indian enterprises with better cost visibility and safer Windows environment automation.
New Usage Breakdown Shows Exactly What’s Consuming Your Credits
Claude Code’s latest update brings much-needed transparency to usage tracking. According to the Claude Code changelog, version 2.1.149 introduces a detailed breakdown in the /usage command that shows exactly what’s driving your API consumption across skills, subagents, plugins, and individual MCP servers.
For Indian businesses managing tight AI budgets, this granular visibility becomes crucial when optimizing costs. A marketing agency in Pune running multiple automated content workflows can now pinpoint whether their social media scheduling agents or their SEO analysis tools are consuming the bulk of their monthly credits. Instead of guessing which automation is expensive, teams get precise per-category breakdowns.
The update also enhances the cost/diff detail view with full keyboard navigation support. You can now scroll through usage reports using arrow keys, j/k shortcuts, Page Up/Down, Space, and Home/End keys—a workflow improvement that speeds up monthly budget reviews significantly.
GitHub-Style Task Lists and Better Visual Feedback
Claude Code 2.1.149 now renders GitHub Flavored Markdown task lists with proper checkboxes instead of plain bullet points. When Claude generates project checklists or to-do items, you’ll see proper - [ ] incomplete task and - [x] completed task formatting that matches GitHub’s visual standards.
This seemingly small change improves collaboration for Indian software teams using Claude Code alongside GitHub workflows. A development team in Bengaluru reviewing sprint planning documents generated by Claude will now see properly formatted task lists that integrate seamlessly with their existing GitHub project management setup.
Critical PowerShell Security Improvements
The most significant fix in this release addresses a PowerShell permission bypass vulnerability. Claude Code’s built-in directory navigation functions (cd.., cd\, cd~, and drive letters like X:) were previously changing working directories without proper detection, allowing subsequent commands to potentially access files outside approved workspace boundaries.
For Indian enterprises using Windows environments with strict data access controls, this security gap posed real risks. A financial services company in Mumbai with sensitive client data could have inadvertently exposed protected directories if Claude Code sessions navigated outside intended project boundaries. The fix ensures all directory changes are properly tracked and enforced within sandbox permissions.
Additionally, the update fixes PowerShell prefix and wildcard permission rules. Previously, rules like PowerShell(dotnet.exe build *) weren’t properly pre-approving matching executables and scripts, causing unnecessary permission prompts during automated builds.
Enterprise Features and MCP Connector Management
Enterprise users gain a new managed setting: allowAllClaudeAiMcps. This allows organizations to load claude.ai cloud MCP connectors alongside their custom managed-mcp.json configurations. Indian enterprises can now blend Anthropic’s curated cloud integrations with their internal tools and databases, creating hybrid automation environments that work with both external services and proprietary systems.
The update also fixes several sandbox and git worktree permission issues that were overly restrictive or incorrectly permissive, ensuring enterprise deployments maintain proper security boundaries while enabling legitimate development workflows.
Performance and Reliability Improvements
Several performance fixes target resource-intensive operations. The find command in the Bash tool previously could exhaust macOS system file limits on large directory trees, potentially crashing the host system. This fix particularly benefits Indian software companies with large codebases or extensive project archives.
The thinking spinner now properly resets between tool calls, eliminating visual confusion where amber “thinking” indicators persisted across multiple operations. For teams monitoring long-running analysis tasks, this provides clearer feedback about Claude’s current processing state.
UI and Navigation Enhancements
Keyboard navigation improvements extend beyond usage reports. The transcript view (Ctrl+O) now properly tails new messages instead of freezing at the opening moment. Status bars correctly show effort levels applied by skills and agents rather than baseline user settings, providing accurate feedback about current processing intensity.
Slash command argument hints no longer clip trailing characters when suggestions overflow input boxes, and Tab-completion works correctly for skills whose frontmatter names differ from directory names—common scenarios when teams customize skill organization.
What This Means for Indian Teams
These improvements directly address pain points experienced by Indian organizations adopting Claude Code for development workflows. The usage transparency helps with budget planning in cost-conscious markets, while PowerShell security fixes remove deployment barriers for Windows-heavy enterprise environments common in Indian IT services companies.
The MCP connector flexibility allows Indian businesses to integrate Claude Code with local systems and compliance requirements while still accessing global cloud services. A Chennai-based fintech startup can now connect Claude to their internal risk assessment APIs while maintaining access to Anthropic’s curated integrations for market data and regulatory updates.
Limitations and Considerations
While this update brings substantial improvements, some limitations remain. The new usage breakdowns require teams to actively monitor the /usage command—there’s no proactive alerting when specific categories approach budget limits. Indian teams managing multiple client projects may need to implement their own monitoring workflows.
The PowerShell security fixes, while necessary, might temporarily disrupt existing automation scripts that relied on the previous permissive behavior. Windows-based teams should test their Claude Code integrations after updating to ensure all workflows continue functioning as expected.
Enterprise MCP connector management requires administrative configuration and may not be immediately available in all deployment environments. Smaller Indian companies using cloud-hosted Claude Code instances should verify feature availability with their deployment setup.
Getting Started with the Update
Claude Code automatically updates in most environments, but you can verify your version by running claude --version in your terminal. The new usage tracking features activate immediately—run /usage to see the detailed breakdown of your current consumption patterns.
For enterprise administrators, review the new allowAllClaudeAiMcps setting documentation to understand integration options with existing MCP configurations. Teams currently experiencing PowerShell permission issues should see immediate improvements, though custom scripts may require testing and adjustment.
Watch for the next releases to address remaining UI edge cases and expand enterprise management capabilities as Anthropic continues rapid development of Claude Code’s enterprise features.
