AI Provider Error Clarity, Builder Favorites, and Template Expansion

October 12, 2025

This release delivers three significant improvements. First, when an AI provider encounters an issue — safety filters, token limits, or service problems — you now get a clear, specific message explaining what happened and what to do about it, with a direct link to the provider’s status page. Second, you can now favorite builders and browse them in multiple view modes. Third, the builder template system has been expanded with comprehensive documentation and support for 40+ builder types across eight categories.

New Features

• Builder favorites — You can now star your most-used builders to keep them easily accessible. A filled yellow star marks your favorites, and a “Show Favorites” filter lets you quickly narrow the list to just your pinned builders. The stats dashboard shows your pinned builder count alongside other metrics.
• Multiple view modes for builders — The builders page now offers four view options: list view, two-column grid, three-column grid, and four-column grid. Choose the density that works best for how you browse — compact list for scanning, or wider grids for visual browsing.
• Sorting options — A new dropdown lets you sort builders alphabetically (A–Z or Z–A), replacing the previous single-button toggle.
• 40+ builder types — The builder template system has been expanded to support over 40 builder types organized into eight categories: Archetypes (8), Compliance (3), Industries (4), Integrations (3), Methodology (4), Personas (2), Specialized (3), and Workflows (3). Each builder type includes a core instruction file and a template examples file.
• Back navigation on detail pages — Both builder and analyzer detail pages now include a “Back to Builders” or “Back to Analyzers” button in the header, making it easy to return to the browsing view after opening a specific builder or analyzer.

Improvements

• Clear, actionable AI error messages — When an AI provider returns an error, you now see a specific explanation tailored to what went wrong. Examples include:
  • Content blocked by the provider’s safety filters — with guidance on how to adjust your prompt
  • Token limit exceeded — with a suggestion to shorten your input or choose a model with a larger context window
  • Incomplete generation — explaining the response was cut short
  • Each error message includes a direct link to the relevant provider’s status page (OpenAI, Anthropic, Google, or xAI) so you can check if the issue is on their end
• Consistent error format across all providers — Error messages follow the same clear structure regardless of which AI provider you’re using, so you always know what the problem is and what action to take.
• Analyzer detail page tips — Analyzer detail pages now include an informational tip box (matching the builder page design) with guidance on how to get the best results from the selected analyzer.
• AI model selector numbering — The model selection section in builders is now labeled “5. AI Model Selection” for consistency with the numbered section pattern used throughout the builder form.