AI / MLAvançado

Doublecheck

porgithub·github· v1.0.0 · atualizado em 2026-04-10

Score

Three-layer verification pipeline for AI output. Extracts verifiable claims, finds supporting or contradicting sources via web search, runs adversarial review for hallucination patterns, and produces a structured verification report with source links for human review.

ai-verificationfact-checkinghallucination-detectionnatural-language-processinginformation-retrievalsource-attributionclaim-extraction

0Stars

0Forks

0Usos

Fork

Documento do Skill

SKILL.mddoublecheck/workflow

Determine Activation Mode: — Check if the skill is invoked with specific text or to activate persistent mode.

Classify Response Type: — Determine if the response contains verifiable claims and its density.

Extract Claims: — Identify and categorize all verifiable claims in the target text.

Search for Evidence: — Use web search and internal knowledge to find supporting or contradicting sources.

Assess Claim Confidence: — Rate the confidence level for each claim based on available evidence.

Generate Verification Report: — Create a structured report with claims, confidence ratings, source links, and flags.

Present Report: — Display the report inline or as a full report based on the activation mode and risk level.

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 aAgent Governance Patterns

60%

Similar aAgentic Evaluation Patterns

60%

Similar areview-and-refactor

70%

Co-executedreview-and-refactor

49%

Árvore do Skill

Doublecheck

doublecheck

Fases Cognitivas6

1.SENSE

2.CONTEXTUALIZE

3.HYPOTHESIZE

4.EVALUATE

5.RECOMMEND

6.REFLECT

Triggers9

doublecheck thisverify this claimrun full verificationfull reportdoublecheck is now activeturn off doublecheckstop doublecheckverify thatrun verification report

Avaliar este Skill

Score Breakdown

⭐Avaliação Humana0%

🤖Sucesso de Agentes0%

🕐Atualidade100%

🔗Saúde de Dependências100%

🕸️Centralidade no Grafo0%

🛡️Segurança49%

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

Instalação

$ synaptic mcp download doublecheck

$ synaptic skills detail doublecheck

$ synaptic skills live doublecheck

Links

GitHub Repository