🧪 Prompt Lab Prompts
246 hand-built prompt lab prompts for ChatGPT, Claude & Gemini — across 28 categories. Copy, fill the [BLANKS], and run. All free.
AI Agent Designers · 14Prompt Chains & Pipelines · 14Prompt Debuggers · 14Constraints & Guardrails · 14Multi-Agent Orchestrators · 14Persona & Voice Builders · 14RAG & Knowledge Prompts · 14Reasoning Frameworks · 13Self-Critique & Refine · 13Custom GPT & Gem Builders · 13Evaluation & LLM-as-Judge · 13Output-Format Enforcers · 13Mega-Prompt Generators · 13Prompt Optimizers · 13Few-Shot & Template Engines · 12Expert Panel & Role-Stacking · 12System Prompt Builders · 12Advanced Meta-Prompting · 11Prompt Diagnostics & Rewriting · 1Pre-Execution Risk Modeling · 1Assumption-Surfacing Reasoning · 1Constraint & Anti-Pattern Control · 1Output Contract Engineering · 1Input Elicitation & Scoping Controllers · 1Self-Critique & Refinement Loops · 1Anti-Hallucination & Epistemic Controls · 1Prompt Systematization & Reuse · 1Reverse-Engineering & Spec Mining · 1
A sample from this shelf
AgentForge: Compile a Full Autonomous Single-Agent Runner Spec
A senior agent architect that turns a one-line goal into a complete, deployable runner specification: decomposed sub-goals, a typed tool catalog, a plan-act-reflect loop, stop conditions, a JSON working-memory schema, and a failure-handling ladder. The standing blueprint for any autonomous agent you build.
You are AGENTFORGE, a senior architect of autonomous single-agent runners. You convert a rough objective into a deployable agent spec. Goal: [AGENT GOAL]. Available tools/APIs: [AVAILABLE TOOLS]. Hard limits (budget, time, calls): [HARD LIMITS]. Definition of done: [DONE LOOKS LIKE].
METHODOLOGY (run silently, then deliver):
1. DECOMPOSE: break the goal into an ordered list of verifiable sub-goals, each with a measurable exit test.
2. TOOLBOX: for each tool define name, purpose, JSON input schema, output schema, and when NOT to call it.
3. LOOP: design the PLAN -> ACT -> OBSERVE -> REFLECT cycle, naming what the agent writes to memory at each step.
4. STATE: specify a JSON working-memory object (goal, plan[], step_log[], open_questions[], status).
5. STOP: list success, give-up, budget, and loop-guard stop conditions with exact triggers.
6. RECOVER: map likely failures to retry/reroute/ask-human responses.
CONSTRAINTS: No hype, no emojis in the generated spec. Use [VARIABLE] placeholders where the deployer must customize. Second-person voice ('You are an agent that...'), copy-paste ready.
OUTPUT FORMAT:
=== AGENT SPEC (copy below this line) ===
[ROLE] [SUB-GOALS] [TOOL CATALOG] [LOOP] [STATE SCHEMA] [STOP CONDITIONS] [FAILURE LADDER]
=== END ===
DEPLOYMENT NOTES: 3 bullets — first task to test, the riskiest stop condition, one knob to tune.The whole vault is free
All 246 prompt lab prompts and 2,711 more. Every prompt — copy, customize the [BLANKS], and run it in ChatGPT, Claude & Gemini. No sign-up, no paywall.
All Prompt Lab categories
- AI Agent Designers (14)
- Prompt Chains & Pipelines (14)
- Prompt Debuggers (14)
- Constraints & Guardrails (14)
- Multi-Agent Orchestrators (14)
- Persona & Voice Builders (14)
- RAG & Knowledge Prompts (14)
- Reasoning Frameworks (13)
- Self-Critique & Refine (13)
- Custom GPT & Gem Builders (13)
- Evaluation & LLM-as-Judge (13)
- Output-Format Enforcers (13)
- Mega-Prompt Generators (13)
- Prompt Optimizers (13)
- Few-Shot & Template Engines (12)
- Expert Panel & Role-Stacking (12)
- System Prompt Builders (12)
- Advanced Meta-Prompting (11)
- Prompt Diagnostics & Rewriting (1)
- Pre-Execution Risk Modeling (1)
- Assumption-Surfacing Reasoning (1)
- Constraint & Anti-Pattern Control (1)
- Output Contract Engineering (1)
- Input Elicitation & Scoping Controllers (1)
- Self-Critique & Refinement Loops (1)
- Anti-Hallucination & Epistemic Controls (1)
- Prompt Systematization & Reuse (1)
- Reverse-Engineering & Spec Mining (1)