note

Belmont: Note

You are saving learnings and discoveries to NOTES.md so they persist across sessions and context compactions. This skill captures non-obvious information — environment quirks, workarounds, debugging insights, credential locations — that would otherwise be lost.

Storage Locations

• Feature-level: {base}/NOTES.md — learnings specific to the current feature
• Global: .belmont/NOTES.md — cross-cutting learnings (env setup, project-wide patterns)

Where {base} is the feature's base path (e.g., .belmont/features/<slug>/).

Step 1: Determine Target

Decide where to save the note:

1. If the user specifies "global" or the note is about project-wide setup/environment, use .belmont/NOTES.md
2. If the user specifies a feature or you're clearly working in a feature context, use {base}/NOTES.md
3. If unclear, ask the user:

Where should this note be saved?
  [1] Feature: <feature-name> ({base}/NOTES.md)
  [2] Global (.belmont/NOTES.md)

Step 2: Extract Learnings

Gather the content to save:

1. If the user provided specific text, use that
2. If the user says "save what we learned" or similar, extract non-obvious discoveries from the conversation:
   • Environment requirements or quirks
   • Workarounds for bugs or limitations
   • Debugging insights that took effort to discover
   • Credential or config file locations (NEVER save actual secret values)
   • Performance findings
   • Non-obvious patterns or conventions

Do NOT save:

• Routine task completion notes (that's what PROGRESS.md is for)
• Obvious information that's in the docs
• Actual secret values, tokens, passwords, or API keys — only save their locations

Step 3: Confirm with User

Show the user what will be saved:

Saving to: {target file path}

### {Category}
- {learning 1}
- {learning 2}

Save this? [y/n]

Wait for confirmation before writing.

Step 4: Write to NOTES.md

1. If the file doesn't exist: Create it with the header and today's entry: