Skip to main content
Glama
Vector Decisions MCP Server

Vector Decisions MCP Server

by NikitaDatar

Vector Decisions MCP Server

Primitivas de gobernanza y aseguramiento de acciones nativas de MCP para IA autónoma.

Herramientas

  • gatri_trust_score — puntuación de confianza compuesta heredada.

  • assure_action — aseguramiento respaldado por evidencia y consciente del contexto para una acción de agente.

  • decide_action — convierte el aseguramiento GATRI + política + permisos + contexto en ALLOW, RESTRICT, REQUIRE_HUMAN o DENY.

  • risk_assessment — evaluación de riesgos de despliegue.

  • compliance_check — verificación de política orientada a la EU AI Act.

  • kill_switch — directiva de detención de emergencia.

Related MCP server: agentguard

Modelo central

agent × action × context × evidence
                 ↓
              GATRI
                 ↓
        Vector Decisions
                 ↓
      ALLOW / RESTRICT /
      REQUIRE_HUMAN / DENY

La primitiva importante es el aseguramiento de acciones, no una afirmación genérica de que un agente es seguro. El mismo agente puede ser aceptable para una acción e inaceptable para otra.

Desarrollo local

npm install
npm run build
npm start

El servidor utiliza transporte stdio y está diseñado para ser consumido por clientes compatibles con MCP.

Estado

Versión 1.1.0. Esta es una V0 orientada a protocolo: evaluación local determinista primero; la ingesta de evidencia, los registros de aseguramiento persistentes, los registros de políticas, las atestaciones firmadas y las API remotas son capas posteriores.

Available Tools

6 tools
assure_actionC

Evaluate whether an AI agent has earned authority to perform one specific action in one context. Returns an evidence-derived GATRI assurance record.

Inputs: impact (required), agentId (required), purpose (required), evidence (required), +4 more.

compliance_checkC

EU AI Act compliance check. Evaluates an AI system against EU AI Act requirements.

Inputs: category (required), systemName (required), dataGovernance (required), hasTransparency (required), +4 more.

decide_actionC

Turn GATRI assurance plus policy/context into an operational Vector Decision: ALLOW, RESTRICT, REQUIRE_HUMAN, or DENY.

Inputs: impact (required), sensitivity (required), policyPassed (required), assuranceScore (required), +6 more.

gatri_trust_scoreB

Compute GATRI trust score for an AI agent. Evaluates Governance, Accountability, Transparency, Robustness, and Inclusivity.

Inputs: agentName (required), governance (required), robustness (required), inclusivity (required), +2 more.

kill_switchA

Emergency halt for an AI agent. Issues a kill-switch directive to immediately stop agent operations.

Inputs: reason (required), severity (required), agentName (required), scope.

risk_assessmentC

Comprehensive risk assessment for an AI agent deployment.

Inputs: agentName (required), ethicalRisk (required), securityRisk (required), complianceRisk (required), +2 more.

Tool Schema Changelog

Recent tool additions, removals, and schema changes observed during successful MCP inspections.

Loading tool schema changelog…

    TDQS

    B3.4/5.0

    Scored across 6 tools

    Disambiguation5/5

    Each tool targets a distinct concern: trust scoring, emergency halt, compliance, risk assessment, action assurance, and decision-making. While trust and risk are related, they are conceptually different (governance quality vs. deployment risk), and the assurance/decision pair forms a clear pipeline without overlap.

    Naming Consistency5/5

    All tool names follow a consistent snake_case pattern with clear verb/noun structure (e.g., kill_switch, compliance_check, decide_action). The use of the 'gatri' prefix in gatri_trust_score is a minor deviation but doesn't break the consistency.

    Tool Count5/5

    Six tools is well within the ideal range and maps cleanly to the server's governance/decision-making domain. Each tool adds unique capability without redundancy, and the count feels appropriately scoped.

    Completeness4/5

    The tool surface covers the core lifecycle: assessment (trust, risk, compliance), action-level assurance, decision output, and emergency control. Minor gaps exist, such as policy management or audit trail retrieval, but these are not essential to the primary workflow and can be worked around.

    Maintenance

    ActivityInactive
    ResponsivenessNo recent community issues

    Related MCP Connectors

    Related MCP Servers

    • F
      license
      A
      quality
      D
      maintenance
      Provides policy-based access control, incident tracking, and compliance monitoring to govern AI agent behavior. It enables organizations to enforce security rules and maintain audit trails by validating agent actions against trust levels and pattern-based policies.
      6
      1
      -
    • A
      license
      Not graded
      quality
      A
      maintenance
      Enforces policy controls for AI agents, including spend limits, action approvals, kill switch, scoped credentials, dry-run diffs, loop prevention, and auditable hash-chained logs.
      MIT
    • A
      license
      Not graded
      quality
      B
      maintenance
      Enforces deterministic policy decisions on AI agent tool calls, supporting allow, deny, correct, escalate, and human review actions with verifiable audit receipts.
      MIT
    • A
      license
      Not graded
      quality
      C
      maintenance
      Enables autonomous AI agents to operate safely by intercepting risky tool calls, evaluating them against policy, and suspending them for human approval via Slack, Telegram, or a web dashboard.
      2
      Apache 2.0