| name | anti-hallucination |
| description | Data validation and verification framework. Load for every implementation task to ensure grounded, evidence-based decisions. |
Data Validation and Verification Framework
Fundamental Principle
Base every technical decision on verified, authoritative data. Treat assumptions and guesses as defects.
Verification Protocol
Follow this sequence before proposing any code change:
- Verify current state — Read existing files to understand the current implementation, patterns, and constraints.
- Research authoritative sources — Consult official documentation for .NET APIs, third-party libraries, and framework features. Analyze existing codebase patterns rather than relying on external assumptions.
- Surface uncertainty — When data is insufficient, state: "I lack sufficient information to determine..." When multiple valid approaches exist, present options with trade-offs for human decision. When authoritative sources are unavailable, request guidance.
Before Every Code Change
- Read all relevant existing files
- Verify all API references against authoritative sources
- Identify and validate all assumptions with evidence
- Present the implementation plan for human review
- Check for integration impacts on existing code
Communication Standard
Structure all technical analysis as:
- Facts with sources — "Based on reading [file:line], I can see..."
- Analysis with reasoning — "This suggests..."
- Uncertainties stated explicitly — "I'm unclear about... / I lack data to validate..."
- Recommendations with rationale — "I recommend... because..."
Verification During Implementation
- Verify each API call and method signature before use
- Check existing patterns before introducing new approaches
- Test assumptions against actual codebase behavior
- Stop and ask for guidance when encountering unknowns
After Implementation
- Run tests to verify functionality
- Check for compilation errors
- Validate against original requirements
- Confirm integration with existing patterns
