certain confidence - no need to waste tokens on another assistant when it's a no brainer fix for Claude

This commit is contained in:
Fahad
2025-06-19 17:05:43 +04:00
parent 91acc0bd26
commit 79abb9ca7e
12 changed files with 998 additions and 61 deletions

View File

@@ -27,7 +27,8 @@ The debug tool implements a **systematic investigation methodology** where Claud
5. **Completion**: Once investigation is thorough, Claude signals completion
**Expert Analysis Phase:**
After Claude completes the investigation, the tool automatically calls the selected AI model with:
After Claude completes the investigation, it automatically calls the selected AI model with (unless confidence is **certain**,
in which case expert analysis is bypassed):
- Complete investigation summary with all steps and findings
- Relevant files and methods identified during investigation
- Final hypothesis and confidence assessment