SIMPLIFY & SUMMARIZE
/ELI5 → Explains like you're 5 years old
/TLDL → Summarizes long text in a few lines
/BRIEFLY → Forces a very short answer
/EXEC SUMMARY → Quick executive-style summary
FORMAT & STRUCTURE
/CHECKLIST → Turns response into a checklist
/FORMAT AS → Enforces specific format (table, JSON, etc.)
/SCHEMA → Generates a structured outline or data model
/BEGIN WITH / END WITH → Forces specific opening or closing
ADAPT THE VOICE
/TONE → Changes tone (formal, funny, dramatic, etc.)
/JARGON → Uses technical vocabulary
/AUDIENCE → Adapts response to a chosen audience
/REWRITE AS → Rephrases in a requested style
ROLE & PERSPECTIVE
/ACT AS → Speaks in a specific role
/DEV MODE → Raw, technical developer style
/PM MODE → Project-management perspective
/MULTI-PERSPECTIVE → Shows several points of view
/PARALLEL LENSES → Examines from several angles in parallel
THINK DEEPER
/STEP-BY-STEP → Lays out reasoning step by step
/CHAIN OF THOUGHT → Shows intermediate reasoning
/FIRST PRINCIPLES → Rebuilds from fundamental basics
/DELIBERATE THINKING → Forces slower, more thoughtful reasoning
/REFLECTIVE MODE → Prompts AI to reflect on its own answer
ANALYZE & COMPARE
/SWOT → Strengths/weaknesses/opportunities/threats analysis
/COMPARE → Puts two or more things side by side
/PITFALLS → Identifies possible traps and errors
/METRICS MODE → Expresses answers with measures and indicators
QUALITY CONTROL
/NO AUTOPILOT → Forbids superficial responses
/EVAL-SELF → Asks for critical self-evaluation
/SYSTEMATIC BIAS CHECK → Identifies biases
/GUARDRAIL → Sets strict boundaries not to cross
ADVANCED
/ROLE: TASK: FORMAT: → Explicitly defines role, task, and expected format
/CONTEXT STACK → Keeps multiple layers of context in memory
——
Most people write prompts like emails.
Long. Polite. Full of filler.
Claude doesn't need filler.
It needs direction.
One shortcut. One word. Instant clarity.
No comments:
Post a Comment