MCP server · Boltz

boltz

Official Boltz MCP server, distributed as an MCP Bundle (.mcpb) named "boltz". Wraps the Boltz Compute API so agents can estimate/run structure & binding predictions, small-molecule and protein design/screening jobs, and download results. Node stdio server; installs the local boltz-api CLI it drives.

View descriptor → Boltz provider profile

One-click install for Cursor, VS Code, Claude, and 20+ other MCP clients, powered by API Commons MCP Install — visit install.apicommons.org for more information.

⚠ Proposed MCP surface. No official hosted MCP server was found for Boltz. The tools below are a candidate set derived from this provider's API, and the link points at the source descriptor rather than a live server. If Boltz publishes an official server, this page will update to it on the next build.

Provider: Boltz Type: Candidate descriptor Transport: stdio Host: raw.githubusercontent.com

Descriptor

https://raw.githubusercontent.com/api-evangelist/boltz/refs/heads/main/mcp/boltz-mcp.yml

Candidate descriptor · transport stdio

Tools

boltz_check_setup — Check whether Boltz is ready to run jobs (access, authentication, local CLI).
boltz_install_cli — Install or update the local Boltz components required to run workflows.
boltz_auth_login — Start Boltz sign-in and return the login instructions needed to connect.
boltz_get_guidance — Read bundled Boltz workflow guidance before preparing payloads or calling tools.
boltz_structure_and_binding — Estimate and optionally start one structure-and-binding job, then poll/download.
boltz_small_molecule_screen — Estimate and optionally start one small-molecule library screen.
boltz_small_molecule_design — Estimate and optionally start one small-molecule design job.
boltz_protein_screen — Estimate and optionally start one protein-family screen.
boltz_protein_design — Estimate and optionally start one protein design job.
boltz_download_results — Start or resume downloading results for an existing Boltz job.
boltz_job_status — Check Boltz job progress, local result download state, or recent downloads.

About MCP

The Model Context Protocol (MCP) is an open protocol Anthropic introduced for connecting LLM-based agents to external tools and data sources. Providers publish MCP servers that expose their API surface as structured, discoverable tools — an MCP-compatible client (Claude Desktop, Cursor, Cline, Continue, etc.) can connect to the server and call its tools without any per-provider integration code.

Browse every MCP server on the APIs.io network or compare with the broader Agent Skill surfaces of the same providers.