Skip to content

Conversation

@kevinold
Copy link

Summary

  • Adds "Customizing Agent Behavior" section to plugin README
  • Provides example CLAUDE.md snippets for LSP prioritization
  • Documents when to use LSP vs grep for code understanding

Context

Instead of modifying plugin agents to prioritize LSP, this approach lets projects opt-in by adding guidance to their own CLAUDE.md. This is cleaner because:

  1. No plugin changes needed
  2. Opt-in for projects that want it
  3. Customizable per project's language/tooling

Test plan

  • Verify README renders correctly on GitHub
  • Test that adding the snippet to a project's CLAUDE.md influences agent behavior

kevinold and others added 2 commits January 15, 2026 13:06
Add "Customizing Agent Behavior" section to README with example
CLAUDE.md snippets that projects can use to prioritize LSP tools
over grep for code understanding.

Includes both minimal (one-line) and detailed versions, plus
guidance on when LSP vs grep is appropriate.

🤖 Generated with [Claude Code](https://claude.com/claude-code)

Co-Authored-By: Claude <noreply@anthropic.com>
- Move documentSymbol to first position with "(ALWAYS START HERE)"
- Add Agent Delegation section with sub-agent instruction
- Add "(MANDATORY)" emphasis to header
- Include Rust in supported languages
- Add sub-agent awareness in introduction

🤖 Generated with [Claude Code](https://claude.com/claude-code)

Co-Authored-By: Claude <noreply@anthropic.com>
@kieranklaassen
Copy link
Collaborator

Thanks! Should we just reach for LSPs instead of grep either way? If so, maybe we can bake it into the main loop? Curious to hear what you think and how it works for you.

@kevinold
Copy link
Author

kevinold commented Jan 21, 2026 via email

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants