LangGuard SCOPE

LangGuard SCOPE is a compliance evaluation tool designed to assess agentic workflows before deployment. It functions as a Claude plugin, providing deterministic pre-flight compliance reports for workflows that involve granting agents access to multiple MCP tools and APIs across platforms such as Salesforce, Stripe, GitHub, Slack, Gmail, payroll systems, observability stacks, and vector stores. The platform addresses the challenge of understanding compliance and risk exposure when agents are empowered to act across organizational boundaries, a scenario that traditional compliance frameworks did not anticipate.

The tool maps every MCP tool and API action to a risk level (low, medium, high, or critical), business impact, and 26 regulatory regimes, including GDPR, HIPAA, PCI, SOX, SOC 2, EU AI Act, ISO 42001, and NY DFS 500. For each action, SCOPE identifies whether it raises segregation-of-duties concerns and provides a concrete recommendation, such as proceeding, proceeding with an audit trail, requiring human review or approval, or blocking and requiring human approval. Reports are generated deterministically from a curated database, ensuring that the same input consistently yields the same output, which enhances auditability and avoids issues common with generative AI outputs like hallucinations or inconsistent classifications.

SCOPE is delivered as a Claude plugin that users can install via the Claude Cowork interface. Access is managed through self-service signup, with API tokens provided upon registration. The default token tier allows up to 10 requests per minute, with higher volumes available upon request. The underlying data used for risk and compliance classification is openly available in a public repository, allowing users to review, challenge, or suggest corrections to specific classifications. A public data catalog is also available, enabling browsing of all classified actions by platform, risk, and regulatory regime without requiring a token.

The platform is intended for organizations and teams building agentic workflows who need to evaluate compliance posture and risk exposure before deploying agents. Its design supports integration into change-management workflows and enables proactive compliance review at the design stage, rather than after deployment.