/context command: MCP tool name wraps incorrectly, merging token count into tool name

Resolved 💬 3 comments Opened Feb 12, 2026 by tarurar Closed Mar 13, 2026

Description

When running /context, long MCP tool names wrap incorrectly in the terminal. The token count gets concatenated with the tail of the tool name instead of appearing as a separate column.

Steps to Reproduce

  1. Have MCP plugins with long tool names configured (e.g., mcp__plugin_microsoft-docs_microsoft-learn__microsoft_code_sample_search)
  2. Run /context
  3. Observe the MCP tools section

Expected Behavior

The tool name and token count should wrap gracefully, keeping the token count visually separated from the tool name.

Actual Behavior

The line:

└ mcp__plugin_microsoft-docs_microsoft-learn__microsoft_code_sample_se365 tokens
rch:

The token count 365 tokens is concatenated directly to the truncated tool name (...se365 tokens), and the remaining rch: is orphaned on the next line. It should read:

└ mcp__plugin_microsoft-docs_microsoft-learn__microsoft_code_sample_search: 365 tokens

Screenshot

<img width="1543" height="470" alt="Image" src="https://github.com/user-attachments/assets/6d10604a-4879-45df-947b-822ee61ea9d9" />

(The token count merges with the end of a long tool name when the line wraps)

Environment

  • Claude Code CLI
  • Terminal with standard width
  • Multiple MCP plugins configured (context7, microsoft-docs, playwright, atlassian)

View original on GitHub ↗

This issue has 3 comments on GitHub. Read the full discussion on GitHub ↗