guard

When this skill is invoked:

1. Call snoopy_list and snoopy_scan (pass the current working directory) in parallel.

2. Present a dashboard like this:

   Blocked patterns: .env, .env.*, *.pem, *.key, ...
   Allowed exceptions: .env.example, .env.template
   
   Blocked files found in project:
     1. .env
     2. config/secrets.json
   
   Exposed (sensitive but not blocked):
     3. deploy/private.key
     4. scripts/token.txt
   
   Allowed files:
     5. .env.example
   

   Number every file across all sections in a single sequence.

3. Ask the user what they want to do. Present these options:

   • Block a file or pattern (add to blocked list)
   • Allow a file or pattern (add to allowed list, overrides blocks)
   • Remove a pattern from blocked or allowed
   • Reset to defaults

4. Let the user pick by number from the list or type a custom glob pattern.

5. After each action, show the updated state briefly.

Tools available:

• snoopy_list — get current blocked and allowed patterns
• snoopy_scan — scan a directory for sensitive files (blocked, exposed, allowed)
• snoopy_block — add a pattern to the blocked list
• snoopy_allow — add a pattern to the allowed list
• snoopy_remove — remove a pattern from either list
• snoopy_reset — restore defaults

Allowed patterns override blocked patterns. Config persists at ~/.snoopy/config.json.