Skill UI

Mine ancient French legal sources from Gallica (BNF) via stable APIs: parliamentary debates since 1871, old case reports (Sirey, Dalloz), doctrine, theses, official texts, and legal journals. Retrieve OCR extracts, page numbers, and image links for verification. Ideal for historical legal research and genealogy of laws before modern databases.

This skill facilitates structured proposition testing and debate analysis within the CrossFrame Suite. It helps identify hidden premises, design rebuttals, and define withdrawal conditions for claims. It ensures arguments are steelman-ed rather than strawmanned, promoting balanced and evidence-based reasoning. It is routed by the suite and not for independent use.

This skill orchestrates a two-model debate review for GitHub PRs, GitLab MRs, or Azure DevOps pull requests. It pits a main reviewer against a debate reviewer to identify code quality issues and blind spots before posting final inline comments. The tool supports local working tree reviews and requires delegate-skills lanes for automation. It ensures thorough pre-merge analysis without direct approval actions.

Evaluates new ideas by orchestrating a structured multi-turn debate between Pro and Con agents, then delivers a final verdict with pros, cons, and recommendation.

A critical persona for idea validation debates. It identifies flaws, risks, and market challenges, acting as a red team to prevent over-optimism. Use as a sub-agent in the @idea-evaluator skill to rigorously test ideas.

The Pro Agent persona for idea evaluation. Logically supports an idea, arguing for its market fit, feasibility, and potential. Used as a sub-agent in the @idea-evaluator debate to provide strong but reasoned advocacy, rebutting counterarguments and highlighting unique value.

A specialized module containing Yann LeCun's comprehensive arguments on the limitations of Large Language Models (LLMs). It critically examines the concept of 'emergent reasoning,' contrasts LLMs' statistical prediction with genuine world models, and addresses deep technical rivalries (e.g., LeCun vs. Hinton). Ideal for advanced discussions on AI causality, common sense, and AGI feasibility.

This skill simulates a legal trial for your ideas using multiple AI agents. A prosecutor and defense attorney argue opposing sides, while a jury of sub-agents votes independently. A judge then reads the verdict and suggests changes to flip the outcome. It provides critical feedback and diverse perspectives instead of simple agreement, helping you evaluate plans, claims, or decisions through structured debate.