Advanced elicitation (sanity,coherence,meta-check, challenge,anti-bias)

[LKrysik]

What

Added 35 advanced elicitation methods to the workflow system, covering anti-bias detection, challenge techniques, meta-checks, sanity validations, and coherence assessments.

Why

Strengthen human-AI collaboration by providing agents with sophisticated self-examination and quality assurance techniques. These methods help catch deception patterns, identify blind spots, verify scope integrity, and ensure new code integrates coherently with existing systems.

How

• Added 4 anti-bias methods (Liar's Trap, Mirror Trap, Confession Paradox, CUI BONO Test) for detecting agent shortcuts and dishonesty
• Added 8 challenge methods based on paradoxes (Barber, Sorites, Newcomb, etc.) to force consideration of alternatives
• Added 7 meta-check methods for self-evaluation limits and 6 sanity checks for scope/alignment/completeness verification
• Added 10 coherence methods to ensure new elements integrate well with existing codebase patterns

Testing

Validated CSV format and structure. Methods follow existing pattern with num,category,method_name,description,output_pattern schema.

[LKrysik]

@bmadcode
I had to create another PR, this time with additional elicitation methods (coherence)
Tested those methods and they are useful for checking overall coherence in (code, docs, any type of result)

[bmadcode]

This is awesome - but I am a bit concerned now to release the file at its current size without some measures to first chunk the csv and optimize the workflow. Is this something you would like to take on?

[LKrysik]

@bmadcode
I will take care about that, will create something lightweight

[alexeyv]

Closing. To be replaced with "something lightweight"