Tool Quality score 34/35 despite all subcategories showing full marks (133/133)

[whiteknightonhorse]

The Tool Quality score displays 34/35 on the releases page, but all four subcategories show 133/133 with green checkmarks:

• Tool descriptions: 133/133 — 12pt
• Parameter descriptions: 133/133 — 11pt
• Annotations: 133/133 — 7pt
• Tool names: 133/133 — 5pt

12 + 11 + 7 + 5 = 35, but the score ring shows 34/35.

There is no indication of which subcategory lost a point or why. The UI shows everything as complete/green.

Steps to Reproduce section:

1. Go to https://smithery.ai/servers/apibase-pro/api-hub/releases
2. Look at the "Tool Quality" section in the latest release score breakdown
3. Observe: all 4 subcategories show green checkmarks and 133/133 completion
4. Observe: the circular score indicator shows 34/35, not 35/35

Expected Behavior section:

If all subcategories are 133/133 (full marks), the Tool Quality score should be 35/35 (12+11+7+5).

Alternatively, if there is a genuine issue reducing the score by 1 point, the UI should indicate which subcategory is affected and why.

Actual Behavior section:

Tool Quality shows 34/35 with no visual indication of where the missing point comes from. All subcategories display green checkmarks and 133/133 completion, which contradicts the total score.

This appears to be either:

1. A rounding/calculation bug in the score aggregation
2. A hidden scoring criterion not displayed in the UI breakdown

Additional Context section:

• Page URL: https://smithery.ai/servers/apibase-pro/api-hub/releases
• Server: apibase-pro/api-hub (133 tools, 3 prompts)
• Browser: Chrome
• OS: macOS
• Note: This server has maintained 100/100 total score previously. All tools have descriptions, .describe() on every Zod parameter, annotations (readOnlyHint/destructiveHint/idempotentHint/openWorldHint), and 3-level dot-notation mcpName

• title fields. We cannot find documentation on what might cause 34/35 when all displayed subcategories are complete.