AgentsAvançado

Testing Prompts With Subagents

porneolabhq·neolabhq· v1.0.0 · atualizado em 2026-04-10

Score

Use when creating or editing any prompt (commands, hooks, skills, subagent instructions) to verify it produces desired behavior. Applies a RED-GREEN-REFACTOR cycle to prompt engineering using subagents for isolated testing.

prompt-engineeringtest-driven-developmentllm-testingsubagent-testingprompt-optimizationred-green-refactorai-agent-testing

Linguagens

TypeScriptJavaScriptPythonJavaC#

0Stars

0Forks

0Usos

Fork

Documento do Skill

SKILL.mdcustomaize-agent-test-prompt/workflow

RED Phase (Baseline Testing): — Design test scenarios appropriate for the prompt type. Launch a subagent WITHOUT the prompt and document its behavior, identifying patterns and severity of failures.

GREEN Phase (Write Minimal Prompt): — Write a prompt addressing the specific baseline failures documented in the RED phase. Focus on conciseness and appropriate degrees of freedom.

Test with Prompt: — Run the same scenarios WITH the prompt using a subagent. Verify that the agent follows instructions, baseline failures no longer occur, and the agent cites the prompt when relevant.

REFACTOR Phase (Optimize Prompt): — Improve the prompt while keeping tests passing. Focus on closing loopholes, improving clarity, reducing tokens, and enhancing structure.

Re-verify: — Test again with a fresh subagent to ensure the prompt still works after refactoring.

Git Commit with Verification: — Commit the verified prompt with a clear message, following specific rules for the commit process.

Telemetria de Agentes

Execuções

total

Taxa de Sucesso

últimos 30d

Latência Média

0.0s

p50

Alucinação

0.0%

detecção

Tokens Entrada

avg 0/exec

Tokens Saída

avg 0/exec

Uso por Plataforma

Skills Relacionados

Similar aSteve Jobs Perspective

60%

Similar aAmazon Working Backwards

60%

Similar aRoundtable Discussion

60%

Árvore do Skill

Testing Prompts With Subagents

customaize-agent-test-prompt

Fases Cognitivas7

1.SENSE

2.CONTEXTUALIZE

3.HYPOTHESIZE

4.EVALUATE

5.RECOMMEND

6.ACT

7.REFLECT

Triggers8

test a promptverify prompt behaviordebug a promptimprove prompt clarityoptimize a promptuse red-green-refactor for promptstest prompt with subagentsbaseline test a prompt

Avaliar este Skill

Score Breakdown

⭐Avaliação Humana0%

🤖Sucesso de Agentes0%

🕐Atualidade100%

🔗Saúde de Dependências100%

🕸️Centralidade no Grafo0%

🛡️Segurança50%

CompositeScore = α·Humano + β·Agente + γ·Recência + δ·Deps + ε·Centralidade + ζ·Segurança

Instalação

$ synaptic mcp download customaize-agent-test-prompt

$ synaptic skills detail customaize-agent-test-prompt

$ synaptic skills live customaize-agent-test-prompt

Links

GitHub Repository