swing-research
๐ฏSkillfrom whynowlab/swing-skills
Swing provides six cognitive firewall skills for Claude Code that counter default AI behaviors like agreeing without evidence, hiding reasoning, and skipping source citations, making AI responses more rigorous and research-backed.
Same repository
whynowlab/swing-skills(6 items)
Installation
npx vibeindex add whynowlab/swing-skills --skill swing-researchnpx skills add whynowlab/swing-skills --skill swing-research~/.claude/skills/swing-research/SKILL.mdSKILL.md
More from this repository5
A cognitive guardrail skill that forces AI agents to ask clarifying questions before acting, preventing assumptions and ensuring agents push back when instructions are ambiguous or incomplete.
Part of Swing, an AI agent guardrails system for Claude Code that provides six cognitive firewalls to prevent AI from defaulting to safe answers, hiding reasoning, agreeing blindly, citing no sources, skipping assumptions, and ignoring failure modes.
A set of six cognitive firewall skills for Claude Code that push AI agents to challenge defaults, cite sources, show reasoning, and verify assumptions instead of taking safe or agreeable shortcuts.
Swing provides AI agent guardrails for Claude Code with 6 firewall skills that enforce safety and quality constraints during agent-driven development. Compatible with Claude Code and Codex CLI, it acts as a protective layer to prevent harmful or low-quality outputs.
Six cognitive firewall skills for Claude Code that push AI agents to cite sources, show reasoning, challenge assumptions, and avoid safe defaults. Addresses six common cognitive failures in AI coding assistants and works with both Claude Code and Codex CLI.