Make Claude your harshest critic: adversarial review and one-shot example workflows
thisdudelikesAI · x · 2026-09-15
Two Claude workflows: Adversarial Review — after finishing a draft, prompt Claude as "the harshest critic in the room" to find every weak argument, then fix what it finds. One-Shot Example — instead of describing what you want, show one perfect example of the output format and say "now do this for X"; examples beat instructions every time.
More from coding & agent
- Codex users report built-in deep-research skill vanishing from CLI overnight — Hot_Independence5160 · 2026-09-15
- deepseek-ai/deepseek-harness hits GitHub with 225k stars: 'Everything is a Plugin' — alex_verem · 2026-09-15
- DeepSeek's open-source dsh harness makes every part of a coding agent a swappable plugin — alex_verem · 2026-09-15
- OSINT MCP Server unifies Shodan, Censys and LinkedIn intelligence in one toolkit — modelcontextprotocol · 2026-09-15
- No-coder builds Raspberry Pi kiosks with Claude, instruction file hits 1288 lines — BloodAngel09 · 2026-09-15
- Weaviate: Piling up chat history makes AI memory worse; Engram reconciles facts instead — philipvollet · 2026-09-15