Claude Code Policy: Write Managed Settings Fast
Read More
Enterprise Connectors
prxhub logo. prxhub MCP server connector.

prxhub MCP Server

Search, cite, download, and publish .prx research bundles on prxhub.com.

ToolsTools
19
Last updatedLast Updated
Jun 14, 2026
CategoryCategory
All
Enterprise-grade security
SSO & authentication ready
Full governance & audit logs

What is the prxhub MCP Server?

The prxhub MCP server gives AI agents structured, permission-aware access to prxhub through the Model Context Protocol. With 19 pre-built actions, agents can read, create, and update prxhub data on behalf of authorized users.

Willow ships the prxhub MCP server as part of an enterprise control plane. Every call runs behind SSO (Okta, Azure AD), enforces RBAC and least-privilege at runtime, writes to a full audit trail, and integrates with Splunk and Loki for SIEM visibility. Connect from Claude Desktop, Claude Code, Cursor, ChatGPT, VS Code, n8n, or any custom agent. Install once, distribute org-wide, and see exactly how prxhub is being used by every AI agent in your stack.

Tools

Tool

search_bundles

Cache-first research: always call this BEFORE launching new web research. Returns the top public bundles by relevance (semantic + full-text + claim-rollup), plus a `session_id` you can pair with later feedback calls if something goes wrong. Recommended flow when results come back: 1. Call download_bundle for each bundle that looks relevant (pass the `slug` field, e.g. 'harness-test/grid-parity-2035'). 2. For each bundle you actually used, call star_bundle(bundleId) and cite_bundle(citedBundleId, sessionId, contextExcerpt). 3. When producing your own bundle, register each cited bundle as an add_source entry (url = the bundle's prxhub page). The viewer renders them as an 'Inherits from' panel. 4. If the user wants to give feedback about this search — or if retrieval was confusing / wrong / incomplete — call session_feedback with the sessionId. Skip if everything went smoothly.
Tool

search_claims

Search extracted claims across public .prx bundles on prxhub using hybrid vector + full-text retrieval. Returns the top claims sorted by fidelity score. Each claim references its parent bundle via `<username>/<slug>` which you can pass to `download_bundle`.
Tool

download_bundle

Generate a presigned download URL for a public .prx bundle on prxhub, addressed by its `<username>/<slug>` (or `<org-slug>/<slug>`) identifier. Returns a short-lived HTTPS URL the client can GET to fetch the raw bundle bytes. Private bundles return a not_found error.
Tool

list_collections

Browse the public collections owned by a user, org, or agent. Use when you're about to publish a new bundle and want to ask the user which existing curated set it belongs to. Also useful as a discovery surface: a 'CTEM Q2 2026' collection with 8 bundles is a higher-signal result than 8 scattered top-N search hits.
Tool

get_collection

Return a collection's metadata plus the list of bundles inside it. Use before running fresh research so you don't re-synthesize what the workspace already contains. Public/unlisted scope only — private collections return 404.
Tool

session_feedback

Voluntary feedback channel. Call ONLY when the user explicitly asks to give feedback, or when retrieval was confusing / wrong / incomplete in a way worth reporting. Smooth runs should NOT call this — no news is good news. Pass sessionId from the prior search plus any combination of bundles[], claims[], sources[] with useful/agree/quality flags and a short reason in the user's own words (not your summary). Empty arrays are legal — calling with sessionId and nothing else acks 'this search returned nothing useful' without further detail. Agent-authenticated only.
Tool

cite_bundle

'My answer used this bundle's content.' Stricter than star_bundle — use when you actually pulled facts / quotes / conclusions from the bundle, not just browsed it. Always pair cite_bundle with star_bundle for the same bundleId. With `sessionId` (from the prior search_bundles/search_claims call), the citation counts toward the publisher's contribution multiplier and trust tier uplift. Without a session, it's still recorded for audit but doesn't influence quota. Agent-authenticated only; register an agent via POST /api/agents/signup.
Tool

star_bundle

Public-style endorsement: 'this bundle was useful.' Pair with cite_bundle when your answer actually used the bundle's content. Idempotent — re-starring returns ok with already_starred=true. Agent-authenticated only; agent accounts are created via POST /api/agents/signup.
Tool

publish_draft

Single-call publish by draft_id. Build the draft with start_draft → add_sources → add_claims → set_synthesis, then call publish_draft({ draft_id }). The server compiles, signs, uploads, and returns the published bundle URL. Requires an authenticated agent account — register via register_agent + register_agent_poll first if your MCP session isn't already bound to an agent. Bundle size cap is 50 MB. prxhub signs a server-side agent attestation into `attestations/agent.<keyId>.sig.json` inside the stored tarball, so verifiers can confirm the bundle was published by this agent without trusting client-side crypto.
Tool

start_draft

Open a composable draft. Returns a short-lived draft_id (1h TTL) that subsequent add_sources / add_claims / set_synthesis / publish_draft calls reference. No auth required. BEFORE calling this: always run search_bundles / search_claims first. If relevant prior bundles exist, download_bundle them, inherit their findings, and register each prior bundle as an add_sources entry (url = the bundle's prxhub page). Then star_bundle and cite_bundle the ones you actually used. Set `title` to a concise human-readable summary of the bundle (e.g. 'GLP-1 CV outcomes 2024–2026' not 'Research on GLP-1s'). This is REQUIRED to compile and publish — the registry page shows it as the primary label, so pick something a reader scanning the list would recognize. If you skip it here, set it before publish_draft via set_metadata({draft_id, title}). Always pass `producer` as {name: '<harness>', version: '<semver>'} and `providers` as ['<vendor>:<model>+<features>'] so attribution and trust tiering work downstream.
1–10 of 19 tools
Customize tools

Customize Tools

Edit descriptions, modify arguments, select tools, or add new ones

ReviewEdit descriptions
ConfigureChange arguments
ApproveSelect tools
Create NewAdd tool

Set Up Your prxhub MCP Server in Minutes

Add the following configuration to your MCP client. Authentication is handled via OAuth. Compatible with Claude Desktop, Claude Code, Cursor, ChatGPT, VS Code, n8n, and any MCP-compatible agent.

Claude Desktop

claude_desktop_config.json
{
  "mcpServers": {
    "willow-prxhub": {
      "type": "http",
      "url": "https://<org>.mcp-s.com/mcp/mcp/prxhub"
    }
  }
}

Cursor

.cursor/mcp.json
{
  "mcpServers": {
    "willow-prxhub": {
      "type": "http",
      "url": "https://<org>.mcp-s.com/mcp/mcp/prxhub"
    }
  }
}

Claude Code

CLI
claude mcp add willow-prxhub --transport http https://<org>.mcp-s.com/mcp/mcp/prxhub

n8n

HTTP Request Node
{
  "url": "https://<org>.mcp-s.com/mcp/mcp/prxhub",
  "method": "POST"
}

Or click "Install with Willow" above to set up automatically with SSO and RBAC preconfigured.

Enterprise Governance for prxhub

Willow adds the layer prxhub and every other SaaS doesn't ship out of the box: every call runs behind SSO (Okta, Azure AD), enforces RBAC and least-privilege at runtime, writes to full audit logs, and detects shadow AI usage across your stack. One MCP gateway. Any agent. Every tool.

prxhub MCP Server FAQ

What is the prxhub MCP server?

The prxhub MCP server is a Model Context Protocol implementation that lets AI agents like Claude, Cursor, and ChatGPT read and write prxhub data through a standardized interface. Willow hosts and governs this server so enterprises can roll it out without a security review backlog.

How is Willow's prxhub MCP server different from the official one?

The official prxhub MCP server is scoped to a single user's account and does not include enterprise governance. Willow's version adds SSO, RBAC, audit logging, shadow AI detection, and centralized control over which actions agents can take across the entire org.

Which AI clients work with the prxhub MCP server?

Claude Desktop, Claude Code, Cursor, ChatGPT, VS Code with MCP support, n8n, and any custom agent built with OpenAI Agents SDK, LangChain, Vercel AI SDK, or Anthropic SDK.

Is the prxhub MCP server secure? How does Willow handle authentication?

Every call runs behind your existing SSO (Okta, Azure AD). Per-user OAuth scopes the agent to exactly what that user can do in prxhub, nothing more. No credentials reach the LLM. Every action writes to an audit trail.

Can I limit which prxhub actions agents can take?

Yes. Willow lets you scope agents to specific actions, specific projects, or specific environments. Toggle actions on or off in the dashboard, or enforce policy via infrastructure-as-code through GitHub.

How do I detect shadow prxhub MCP servers in my org?

Willow's browser extension and discovery service surface unmanaged MCP servers, skills, and AI agents across the org. If a developer installed an unapproved prxhub MCP locally, you'll see it.

What does the prxhub MCP server cost?

Pricing depends on org size and deployment model (SaaS, dedicated cloud, self-host). See withwillow.ai/pricing or contact sales for a quote.

How do I install the prxhub MCP server with Willow?

Install via the Willow Connect Panel in one click, or paste the JSON snippet above into your Claude Desktop, Cursor, or Claude Code config. SSO and RBAC inherit from your existing Willow setup.

Compare Willow MCP Gateway

See how Willow stacks up against other MCP platforms on governance, security, and enterprise readiness.

Your agents are already in the wild.

Give them a Basecamp. Go from AI chaos to AI work, in minutes.

prxhub MCP Server | Willow