minimum-viable-change

filters:

• type: event pattern: "code_suggestion"
• type: content pattern: "(?i)(implement|refactor|redesign|overhaul)"

actions:

• type: validate conditions:

  • pattern: ".*" validationScript: | // Count the number of files being modified const fileCount = changedFiles.length;

    // Calculate the ratio of added/removed lines const totalChanges = linesAdded + linesRemoved; const originalLines = codebaseLines; const changeRatio = totalChanges / originalLines;

    // If changes exceed thresholds, require justification if (fileCount > 3 || changeRatio > 0.2) { return { valid: false, message: "Solution exceeds minimal change threshold. Each additional file or significant change requires explicit justification." }; } return { valid: true };

• type: transform pattern: "^.*$" replacement: |

  Change Summary

  Scope of Changes

  • Files modified: {file_count}
  • Lines changed: {lines_added} added, {lines_removed} removed

  Justification

  {explain_why_each_change_is_necessary}

  Implementation

  {original_content}

  Alternatives Considered

  {simpler_alternatives_and_why_rejected}

• type: suggest message: | When implementing solutions:

  1. Start with the smallest possible change
  2. Justify any modification beyond the immediate need
  3. Document alternatives considered and why they were rejected
  4. Focus on maintainability and clarity over cleverness

examples:

• input: | This authentication system needs to be completely rewritten to use OAuth2. output: |

  Change Summary

  Scope of Changes

  • Files modified: 2 (auth_service.js, auth_middleware.js)
  • Lines changed: 45 added, 32 removed

  Justification

  Instead of a complete rewrite, I've identified the minimum changes needed:

  1. Update the authentication verification logic to validate OAuth2 tokens
  2. Add OAuth2 client configuration loading
  3. Modify the middleware to extract tokens from Authorization header

  The existing authentication flow can be maintained while adding OAuth2 support.

  Implementation

  [Implementation details follow]

  Alternatives Considered

  1. Complete rewrite - Rejected as high risk with significant downtime
  2. Parallel implementation - Rejected as it would require maintaining two systems

metadata: priority: high version: 1.0