About Assay Truth Graph MCP
Assay Truth Graph MCP is a remote MCP server for go-to-market teams, providing agents with governed company context and cited evidence. It combines read-only guidance and validation with human-approved write workflows.
Key features
- 01Governed company context and evidence
- 02Audience-specific briefs and proof points
- 03Draft checks against approved claims
- 04Truth Graph search and updates
- 05Collateral and readiness workflows
Use cases
- 01Prepare for a customer meeting
- 02Check a sales draft before sending
- 03Review claims and supporting evidence
- 04Update company records with approval
MCP support
Exposes Assay’s Truth Graph and GTM workflows to MCP clients through a remote streamable HTTP endpoint. Read and verification tools are separate from write and run tools, which require a human-confirmed plan token.
Configuration
Remote servers need no local install. Pick your client, then copy the config or install in one click.
https://app.assay.wiki/api/mcp/v2/mcpclaude mcp add --transport http assay-org-assay-truth-graph https://app.assay.wiki/api/mcp/v2/mcp{"mcpServers":{"assay-org-assay-truth-graph":{"type":"http","url":"https://app.assay.wiki/api/mcp/v2/mcp"}}}Open in Cursor{"servers":{"assay-org-assay-truth-graph":{"type":"http","url":"https://app.assay.wiki/api/mcp/v2/mcp"}}}Open in VS Code{"mcpServers":{"assay-org-assay-truth-graph":{"type":"http","url":"https://app.assay.wiki/api/mcp/v2/mcp"}}}{"mcpServers":{"assay-org-assay-truth-graph":{"serverUrl":"https://app.assay.wiki/api/mcp/v2/mcp"}}}Run commands in a terminal; paste JSON into the client’s MCP config and restart. Servers that need sign-in will prompt on first connect.
Tools
Details
From the project READMEAssay Truth Graph MCP
Assay Truth Graph MCP is a remote MCP server for GTM teams that need AI agents to work from governed, cited company truth.
Assay turns positioning, proof points, brand voice, personas, collateral context, and validation rules into a governed Truth Graph. The remote MCP server hands that context to Claude, ChatGPT, Cursor, Claude Code, Codex, and other MCP clients without copying company facts into each agent.
Two tools are the front door, and most work needs nothing else:
assay(read-only): send the user's request and what you know (an account, people's titles, the channel). You get what to lead with and whether a named person approved it, the approved proof to cite, objections, claims to avoid, what changed since last time, and what is missing with a link to fix it. It also checks a draft before it ships, and answers questions about rivals.assay_act(writes, always two-step): record what happened (an objection, a deal outcome, a message that landed), keep a thread for an account or deal, confirm who a situation is about, or have Assay draft something. Every action previews first and runs only with the plan token after the person agrees.
Behind them, the server has full parity with the Assay app: Truth Graph search
and governed mutation, content validation, collateral, AI search and GEO,
readiness simulation, pre-call briefs, and review and promotion. Call
tools/list on the live endpoint for the current surface.
Connect
Use the remote streamable HTTP endpoint:
https://app.assay.wiki/api/mcp/v2/mcp
Claude Code:
claude mcp add --transport http assay-truth-graph https://app.assay.wiki/api/mcp/v2/mcp
Generic MCP config:
{
"mcpServers": {
"assay-truth-graph": {
"type": "http",
"url": "https://app.assay.wiki/api/mcp/v2/mcp"
}
}
}
OAuth opens in the browser on first use for interactive clients. Headless agents can authenticate with an org API key from Assay Settings.
Replit one-click install:
Codex repo marketplace:
codex plugin marketplace add Assay-Org/assay-truth-graph-mcp
Then install the Assay Truth Graph plugin from Codex Plugins.
What Agents Can Do
- Get a brief for a meeting, deal, launch or board conversation: what to lead with for this audience, who approved it, and the proof to cite.
- Prepare for several accounts in one call.
- Get verified GTM context for customer-facing drafts.
- Search the Truth Graph with provenance and caller-aware context shaping.
- Validate draft content against approved facts and guardrails before it ships.
- Create or update Truth Graph records through dry-run and human-approved plan-token workflows.
- Generate grounded collateral drafts in quarantine, with preview and export gates, and re-render them when the underlying story changes.
- Run the AI search and GEO workspace: briefs, artifacts, evidence, goals, hosted sites, and approval-gated publishing.
- Run readiness simulations and build pre-call briefs.
- Review, approve, and promote through the same gates the app uses.
Two entitlement lanes are enforced server-side. mcp_access covers read and
verification tools on Free and above. mcp_control_access covers write and run
tools on Scale and above. Writes are dry-run by default and mint a single-use,
expiring plan token that must be confirmed. Pricing, policy, proof, and identity
kinds are refused for autonomous mutation, and unmapped kinds fail closed.
Links
- Product: https://assay.wiki/
- MCP docs: https://assay.wiki/mcp/
- Trust center: https://assay.wiki/trust/
- Privacy: https://assay.wiki/legal/privacy/
- Settings: https://app.assay.wiki/settings
- Manual listing checklist: ./MANUAL-SUBMISSION-CHECKLIST.md
- Directory tracker: ./DIRECTORY-SUBMISSIONS.md
Directory Copy
Short description (standard field):
Ranked, governed, cited GTM context for your agents, on one living source of truth.
Short description (hard short limits, 47 chars):
Governed, cited GTM context for your AI agents.
Company one-liner (about-the-company slots):
Elite AI GTM team on one living source of truth.
Tags:
MCP, Model Context Protocol, remote MCP server, GTM, go-to-market, marketing,
sales, content, SEO, brand, positioning, messaging, sales enablement,
content governance, AI governance, compliance, truth graph, provenance,
knowledge graph, Claude, ChatGPT, Cursor
Comments
New to MCP?
From what the protocol is to installing your first server, these answers cover most first questions.
An open protocol that defines how AI applications connect to external tools and data in a uniform way. Clients talk to the model; MCP servers provide tools, resources and prompt templates. They exchange standard messages, so one server can be reused by many clients.