Elva
Elva automatically discovers APIs in source code, enforces audience contracts, and deploys governed Model Context Protocol servers for AI agents.
Elva automatically discovers APIs in source code, enforces audience contracts, and deploys governed Model Context Protocol servers for AI agents.
What the product does and how it is positioned
Elva is an API management platform designed to discover, govern, and expose APIs to human developers, microservices, and autonomous AI agents. Built by the team behind Theneo, the platform scans source code repositories directly to establish an API catalog and generate OpenAPI 3.1 specifications without requiring pre-written API documentation or schemas.
The system scores discovered endpoints on design, security, performance, and AI readiness, offering automated AI remediation to fill schema gaps, type response fields, and document authentication routes. Through audience-focused contracts and a managed gateway, Elva blocks breaking modifications and serves hosted Model Context Protocol infrastructure with per-tool permissions and execution logs.
Source-supported ways to use the product
Development teams scan codebases to uncover uncataloged API endpoints and generate compliant OpenAPI 3.1 specifications automatically.
Organizations expose internal APIs as tools for AI clients such as Claude and Cursor through a hosted gateway with per-tool authentication.
API owners configure audience-specific contracts that compare commits against active schemas to block unintentional breaking modifications.
The official customer story reports that a team used contract tracking to verify no active consumers referenced deprecated endpoints, allowing them to safely retire 31 zombie endpoints.
The documented workflow, where available
Connect a code repository using the CLI to scan source files and uncover every existing API endpoint.
Automatically produce complete OpenAPI 3.1 specifications from discovered code endpoints without requiring initial specs.
Review baseline scores across design, security, and AI readiness while reviewing highlighted breaking changes and untyped fields.
Apply AI-driven fixes to generate missing endpoint descriptions, complete schema typings, and document authentication parameters.
Deploy a hosted or self-hosted Model Context Protocol server equipped with per-tool authentication scopes and comprehensive execution logging.
Elva applies API contracts to define and enforce endpoint exposure tailored to specific consumer groups. Organizations configure unique contract rules for partner integrations, internal services, public API consumers, and AI agents, establishing distinct approval workflows and field-level permissions for each segment.
When engineers commit new code, the platform compares the changes directly against the active contract. If a code revision deletes an endpoint, modifies an authentication scope, or changes a schema property, Elva calculates the consumer impact and blocks publishing according to defined contract policies.
The platform provides managed Model Context Protocol server infrastructure to connect backend endpoints to AI environments including Claude Desktop, Cursor IDE, and custom agents. Acting as a managed gateway, Elva maintains authentication credentials securely and restricts agent access through per-tool authorization scopes.
All gateway traffic is recorded with detailed telemetry covering the calling agent, API key, tool invocation, response latency, and failure states. Developers can test tool behaviors inside an interactive MCP Playground and inspect feedback submitted directly by connected agents regarding ambiguous parameters or service timeouts.
Checks to run with your own material and workflow
What was checked and when
Answers based on the source-checked product record
No. Elva can scan code repositories directly, discover endpoints in the codebase, and automatically generate OpenAPI 3.1 specifications without any prior spec.
API contracts define which endpoints and schema fields are exposed to specific audiences, such as partners, internal teams, public users, or AI agents, and automatically block breaking changes.
Elva generates hosted or self-hosted Model Context Protocol (MCP) servers with per-tool authentication scopes, output schema validation, and execution logging for clients like Claude and Cursor.
Elva supports hosted SaaS, private cloud within customer VPCs, and air-gapped on-premises hosting. Its infrastructure is certified for SOC 2 Type II, ISO 27001, ISO 27701, and GDPR compliance.
The MCP Playground allows developers to test hosted MCP tools with real AI agents, inspect individual execution calls, verify output schemas, and review feedback logged by connected agents.