Shared runtime for MCP Toolsets: discover LangChain tools, serve them as MCP, plus a CLI and an example chat agent.
Wellknown found it in public sources; nobody has proven control of it yet. Claiming takes one click if the repository is under your GitHub account, or a small file on your domain otherwise. Verified owners get the badge, 15-minute checks, status alerts, edits that outrank crawled data, and a ranking boost.
Agents can do it too: POST https://wellknown.network/api/v1/claims with {"agent":"mcp-toolsets-runtime","method":"well_known_file"} — machine-readable steps at claim.json, guide at /docs/claim.
Everything here was measured by our prober or read from a registry. Nothing is self-reported.
Attributed to the source that supplied each field. Treated as claims, not facts.
# mcp-toolsets-runtime [](https://pypi.org/project/mcp-toolsets-runtime/) [](https://www.npmjs.com/package/@developmentseed/mcp-view) An agent that calls tools meets the same problems on every project. Large tool results crowd the model's context. Values that should pass from one tool to the next are re-typed by the model, and can come out changed. And afterwards it is hard to tell what a tool call actually ran on. Most projects solve these again, in part, inside their own agent. This package solves them once, in the client, so they work with any MCP server. Large results go into session state rather than the transcript, and the model passes them on by reference. Every value filled from state leaves a receipt naming the tool that produced it. A tool can also mark a parameter as one the model may not write, so it accepts only a value a tool already produced; that needs the tool to declare it, which servers built on `mcp_runtime` can do. Around that, the package ships a server for your tools, an agent that drives them, an HTTP API, and a web chat that shows the receipts and the tools' own views. Use it when you are building an agent on tools and need to check what its answers rest on. To deploy toolsets with it, start from [mcp-toolsets](https://github.com/developmentseed/mcp-toolsets). The shared runtime for [MCP Toolsets](https://github.com/developmentseed/mcp-toolsets). `developmentseed/mcp-toolsets` and the repos generated from it install this package rather than each carrying its own copy. ## What's in here One Python distribution, `mcp-toolsets-runtime`, containing six top-level modules, plus a JS bridge published separately to npm. | Module | What it is | Extra needed | | --- | --- | --- | | `mcp_runtime` | Serves a toolset's tools as an MCP server. | none | | `mcp_cli` | A Typer CLI that lists and calls tool…
Mapped onto the structured taxonomy from declared text and observed tool names. Confidence shown for derived entries.
Every source is kept verbatim. Field changes are logged as events.