Show HN: Code and reviews that only count if they survive an attack

Category: devtools

Tags: llm-code-generation, adversarial-verification, claude-code

Score: 6.3/10 (Innovation: 6, Technical: 7, Documentation: 7, Utility: 5)

Pantheon-skills provides a multi-agent harness for Claude Code that runs coding tasks through parallel implementations, adversarial verification, and a judge, improving correctness on testable work. Its inclusion of cross-model verification with GPT-5.5 and a separate gap analysis tool makes it a novel orchestration layer for LLM-based development. While the project repackages known techniques, it packages them into a cohesive workflow that addresses a real need for reliable code generation.

Target audience: backend devs, devops, data engineers

Repository: https://github.com/lolu1032/pantheon-skills · JavaScript · MIT · 1 stars

View on Hacker News