Six cognitive failures. Six firewalls. Make it swing.
npx skills add whynowlab/swing-skills --all
The baseline gave a wrong answer confidently.
The baseline found nothing. The skill found a Critical vulnerability.
Premature Closure. Structured clarification before execution.
Hallucination. Source-traced claims with S/A/B/C tier grading.
Confirmation Bias. Steel-man then 3-vector attack.
Anchoring Bias. 5 probability-weighted options.
Black-box Reasoning. Assumption inventory + weakest-link.
Optimism Bias. 5 failure scenarios with circuit breakers.
| Tier | Skills | Use Case |
|---|---|---|
| Daily Use | swing-clarify, swing-research, swing-review | Standard development, PR reviews, architectural questions. |
| High-Stakes | swing-options, swing-mortem | Database migrations, auth flows, payment gateways. |
| Deep Analysis | swing-trace, swing-trace --full | Debugging complex state, legacy code archeology. |
Injects rules only when invoked, saving token context for your actual codebase instead of bloating system prompts.
Chain skills together like Unix pipes. Use trace to find the bug, options to fix it, and mortem to break the fix.
Works across CLI, IDE extensions, and standard chat interfaces. Your cognitive firewalls travel with you.
Open-source prompts. Battle-tested against real-world hallucinations and logic gaps by senior engineers.
npx skills add whynowlab/swing-skills --all
npx skills add whynowlab/swing-skills/swing-review
git clone https://github.com/whynowlab/swing-skills.git && cp -r swing-skills ~/.skills/
cp -r ~/.skills/swing-* ./.claude/skills/