Selqor Labs product — interface layer for agentic systems
Selqor Forge is the MCP builder from Selqor Labs. It reads OpenAPI and Swagger specs, groups APIs by agent intent to keep context minimal, scores quality, and generates TypeScript or Rust MCP server targets — with a no-code visual interface for human refinement and built-in security scanning for hardened deployments.
Explore a sample integration to compare raw API surface area with Selqor Forge's curated tool plan.
Most API-to-tool flows dump endpoints directly into the agent. Selqor Forge analyzes the surface, groups related operations, and keeps coverage while reducing tool sprawl.
Selqor Forge groups related pet, order, and user operations into a smaller MCP surface without losing endpoint coverage.
https://petstore3.swagger.io/api/v3/openapi.json
Curated tools preserve capability while reducing context overhead and tool-selection noise.
MCP is becoming the standard interface for agentic systems. The bottleneck is no longer whether agents can call tools, but how quickly teams can turn raw API specs into usable, governed, high-quality tool surfaces.
Today's MCP generators create one tool per API endpoint — a 50-endpoint spec produces 50 tools. Research shows models lose up to 85% accuracy in large tool spaces, and tool definitions alone can consume over 70% of the context window. The bottleneck is not generation — it is curation.
Even after generation, platform teams still have to wire auth, validate provider access, review outputs, and prepare runtime commands for real deployment.
Basic converters do not tell you whether the resulting tool surface is coherent, compressed enough, or complete enough to work well in agent runtimes.
Read specs from local files or HTTP(S) URLs, extract endpoints, parameters, schemas, security schemes, and normalize them into one consistent model.
Forge maps endpoints into domains and intents so tool planning is based on what the API means, not just how the raw paths are arranged.
Forge groups APIs by what the target agent needs to accomplish — not by endpoint structure. This keeps context windows minimal and tool selection sharp, solving the biggest performance killer in MCP servers today.
Forge measures coverage, compression, description clarity, and schema completeness, and can benchmark curated plans against endpoint-mirror baselines.
Visualize and refine tool groupings, manage auth profiles, test connections, and prepare deployments — all through a no-code interface. Take over and adjust any grouping in minutes when you need human precision.
Generate MCP server targets for different stacks and transports, then prepare deployment commands and environment files for the runtime path you need.
Point Forge at any MCP server URL and run a vulnerability scan. Get a detailed report of security risks, misconfigurations, and exposure issues — so you can harden your deployment before agents go live.
Forge optimizes for the smallest possible tool surface per agent use case. Fewer tools, tighter schemas, and intent-aligned grouping mean agents spend tokens on doing work — not parsing irrelevant context.
openapi-mcp-generator, Higress, FastMCP, AWS Labs, fastapi_mcp — all create one tool per API endpoint. A 50-endpoint spec produces 50 tools. None of them group endpoints into fewer, intent-aligned tools. Forge is the only converter built to solve the context problem, not just the generation problem.
Turn scattered internal OpenAPI specs into one repeatable MCP workflow so teams stop hand-building bespoke tool adapters across the organization.
Expose a curated tool surface for customers building agents without building and maintaining separate integrations for each host or runtime.
Use the dashboard and generated artifacts to review outputs, test auth profiles, manage providers, and prepare deployable runtime commands.
Compare curated plans against endpoint-mirror baselines and iterate on tool quality instead of treating MCP generation as a one-shot export step.
Supports OpenAPI 3.x and Swagger 2.0. Requires Rust 1.85+ and Node.js 20+ for generated TypeScript targets.
Selqor Forge is a CLI and visual dashboard product by Selqor Labs that turns OpenAPI and Swagger specs into curated MCP server targets. It groups APIs by agent intent to keep context minimal, scores quality, provides a no-code interface for human refinement, and includes MCP server security scanning to harden deployments before agents go live.
MCP (Model Context Protocol) is an open standard created by Anthropic for connecting AI agents to external tools and data sources. It is supported by Claude, ChatGPT, Gemini, GitHub Copilot, and many other AI platforms. MCP lets AI agents call APIs, query databases, and interact with external services in a standardized way.
Yes. Selqor Forge is open source under the Apache 2.0 license. The code is publicly available on GitHub and free to use commercially, with the standard patent grant and attribution requirements of Apache 2.0.
Selqor Forge supports OpenAPI 3.x and Swagger 2.0 specifications in YAML or JSON. It parses endpoints, parameters, request and response schemas, and security schemes from local files or HTTP(S) URLs.
Most converters stop at raw export. Selqor Forge adds semantic normalization, curated tool grouping, measurable quality scoring, benchmark workflows, TypeScript and Rust targets, and a dashboard for auth profiles, provider configs, artifacts, and deployment preparation.
GroundQA is the Selqor Labs reliability layer — an AI-native unified QA platform that unifies AI agent testing, web app testing, and scale testing in one workspace. It delivers 10X QA team productivity through evidence-backed test generation, drift analysis, QA memory, and continuous learning. Learn more at groundqa.ai.
Build the interface layer
agents depend on.
Explore the repo, generate a run, and use Forge to standardize how agents access the APIs you already operate.