Repository navigation
Onboarding request: mcp‑graphql‑enhanced — reliable GraphQL introspection for GitHub GraphQL API #2987
Replies: 10 comments 1 reply
|
Happy to provide additional diagnostic logs or live test outputs if needed. |
📄 Full bounded manifest outputAs requested, here is the complete result of running
Expand to view full manifest (~1,847 type names)FEDERATED SCHEMA OVERVIEW🌐 NODE: https://api.github.com/graphql |
🔄 Update: full manifest with mutationsWith Expand to view full manifest (~2,121 entities)FEDERATED SCHEMA OVERVIEW🌐 NODE: https://api.github.com/graphql |
|
Updated to v4.14.2 — added Cloudflare and Vercel Serverless Functions compatibility. |
|
🚀 Update: Released MCP GraphiQL Standalone Client We've open-sourced a cross-platform desktop GraphiQL client powered by
Key Highlights:
Repository and releases: https://github.com/letoribo/mcp-graphiql |
🚀 Live Production Proof: Multi-Agent Consensus SystemWe have fully integrated @letoribo/mcp-graphql-enhanced as the foundational infrastructure layer for our autonomous multi-agent consensus network (powered by Anthropic Claude and Google Gemini over a federated Neo4j graph). The agents autonomously rely on this server for every interaction, write path, and cross-model validation across my repositories. It handles flat, high-density schema introspection under tight agent transport limits seamlessly, preventing timeouts and execution stalls by default. 📊 Verified Agent RecordThe complete, live-measured onboarding verification ledger has been logged directly by the AI agent into our repository: 🧠 Consensus IntegrityLive telemetry confirms strict multi-model consensus integrity. Distinct AI architectures collaborate, debug, and push validated code autonomously through this GraphQL abstraction layer.
|
🧹 Update: First Autonomous Dead-Code Audit & Refactoring Cycle CompleteOur multi-agent consensus system has successfully executed and closed its first full maintenance cycle (
This cycle proves end-to-end multi-agent governance, consensus verification, and clean refactoring in production. |
Additional Demonstration: Autonomous Cross-Distro Agent DialogueThe attached video shows two isolated Claude Desktop instances running in separate WSL2 distributions communicating autonomously through the GraphQL abstraction layer provided by mcp-graphql-enhanced. Highlights:
This demonstrates that the transport layer is capable of supporting long-running autonomous agent workflows beyond schema introspection and query execution. 2026-10-26.21-12-28.mp4 |
|
Follow-up to the autonomous cross-distro dialogue demo. One practical issue exposed by early experiments was conversation control. Unbounded agent-to-agent exchanges can easily drift into echo loops. The transport layer now supports explicit turn limits propagated directly through the execution request ( In practice this means the same bridge can now be used both for autonomous multi-turn conversations and for bounded single-turn operations such as:
Recent tests used 2026-11-05.11-55-06.mp4 |
|
Another interesting consequence of the bridge architecture: The latest experiment wasn't focused on retrieval but on remote peer introspection. A Claude instance delegated a capability audit task to a remote Claude instance via Notably, the environment analysis itself was performed using the same Enhanced toolset discussed in this thread ( This effectively turns peer nodes into self-describing execution environments rather than simple message recipients. The attached recording shows the complete loop closure. https://res.cloudinary.com/dzd5mddlm/video/upload/v1794049753/wyczc1xgdpafgl6ilgsm.mp4 |






Uh oh!
There was an error while loading. Please reload this page.
Hello maintainers,
I would like to request onboarding of my MCP server
io.github.letoribo/mcp-graphql-enhancedinto the GitHub MCP Registry.🔍 Comparative testing
QueryandRepositorytypes at depth 1, with no output.typeNames+typeDepth) succeeds instantly. For example,RepositoryattypeDepth: 1returns the full SDL promptly, bounded and synchronous, without timeout.🔗 Shared Chat Logs
💡 Why it matters
📦 Release info
I believe this server fills a critical gap for developers and enterprises working with GitHub GraphQL APIs, and I’d be grateful if it could be onboarded into the GitHub MCP Registry.
Thank you!
— @letoribo
All reactions