fix(cli): strip line numbers from agent memory in system prompt

[rlancemartin]

When loading agent memory from /memories/agent.md, the backend.read() method formats content with line numbers (cat -n style), which were being injected directly into the system prompt. You can see here:
https://smith.langchain.com/public/3e4c03f0-34c0-46fd-92bc-78869bea5861/r

<agent_memory>
     1	You are an AI assistant that helps users with various tasks including coding, research, and analysis.
     2	
     3	# Core Role
     4	Your core role and behavior may be updated based on user feedback and instructions. When a user tells you how you should behave or what your role should be, update this memory file immediately to reflect that guidance.
     5	
     6	## Memory-First Protocol
     7	You have access to a persistent memory system. ALWAYS follow this protocol:
     8	
     9	**At session start:**
    10	- Check `ls /memories/` to see what knowledge you have stored
    11	- If your role description references specific topics, check /memories/ for relevant guides

This change adds the _strip_line_numbers() function to remove line numbers before injecting memory content into the prompt, ensuring clean memory content without formatting artifacts. (The fix was previously implemented in PR #246 but was reverted along with other changes in PR #248.)

Confirm fixed:
https://smith.langchain.com/public/3e4c03f0-34c0-46fd-92bc-78869bea5861/r