Put your governed GTM knowledge inside every AI tool. Spekit MCP connects Claude, ChatGPT, Copilot, and Glean to your single source of truth, and brings what your team creates there back into Spekit. Everything reps ask, draft, and send is accurate, on-brand, and current.
Power your entire GTM stack with a single, governed knowledge engine. When your AI tools pull from one governed source, every answer and asset they produce is trusted, accurate, and on-brand.
Automate product launch kits
Instantly generate comprehensive launch kits and enablement materials from your existing product documentation, ensuring your team is ready on day one and push back to Spekit effortlessly.
Unified search & knowledge retrieval
Stop tab-switching. Search and retrieve verified and governed GTM knowledge directly inside any LLM to answer customer queries with accuracy.
Bulk content updates
Rapidly update content at scale by fetching all resources on a topic and updating them in one go using Spekit's MCP.
Deal Room & content creation
Generate Deal Rooms with a single prompt. Pull from your governed source of truth, add the context already in your AI conversation, and publish deal-ready content straight into the room.
Spekit MCP is built on the GTM Knowledge Engine: the governed source of truth that keeps every GTM resource accurate and current. Connect it once, and every AI tool in your stack answers from the same approved knowledge. No more agents citing last quarter's pricing or using outdated decks.
Layers of context wrapped around every piece of knowledge that make it AI-ready. A Spek is a knowledge container with metadata, sources, governance, and access controls wrapped around it - designed to be reusable across any GTM need.
Generic AI assistants answer from whatever they can find: stale decks, old wikis, or the open web. Spekit MCP grounds every AI tool in knowledge that is governed at the source, so reps and agents both work from what your team has actually approved.
One governed knowledge layer. Trustworthy AI for your entire revenue org.
Most teams are running AI that's confidently making things up. Spekit MCP grounds every draft, email, sequence, and message in verified truth, in Claude, ChatGPT, Copilot, Glean, or anywhere else your reps work.
Create knowledge, custom one-pagers, and deal room content in any AI tool, and the output comes back on-brand, accurate, and with approved messaging pushed directly into Spekit.
Stop switching tabs to manage your pipeline. Create Deal Rooms and manage your GTM content directly from Claude or ChatGPT, so your workflow stays seamless and your deals keep moving.
Trusted by revenue teams that scale. See all customer stories ->
Connect through ChatGPT, Codex, or custom agents
Access and update Spekit through Claude.ai, Claude Code, CoWork, or custom agents
Connect to Spekit through Gemini and related tools
Spekit MCP is an implementation of the Model Context Protocol that connects Spekit's governed GTM knowledge to external AI tools. It lets Claude, ChatGPT, Copilot, and Glean retrieve accurate, role-appropriate knowledge from Spekit and push content back, so the answers and assets those tools produce stay grounded in your approved GTM source.
A standard API integration connects two specific systems, usually in one direction. MCP is an open protocol built for AI agents: any compatible AI tool can retrieve from Spekit and push content back without a custom connector for each tool. One MCP server in Spekit reaches every MCP-compatible host in your stack.
Spekit works alongside the AI tools your team already uses, not instead of them. Through MCP, Spekit connects its governed GTM knowledge to Claude, ChatGPT, Copilot, and Glean, so the answers those tools produce are grounded in approved content. You choose the LLM and the host. Spekit is the knowledge layer underneath.
Spekit MCP grounds AI-generated answers in content authored, approved, and governed inside Spekit. Instead of retrieving from broad training data, the AI tool retrieves from your playbooks, pricing, messaging, and competitive intel, kept current through Spekit's governance controls. The result is accurate answers rather than confident guesses.
Spekit is the single governed source. When pricing, messaging, or positioning changes, the update happens once in Spekit and flows through MCP to every connected AI tool. Reps working in Claude and reps working in ChatGPT both retrieve from the same current, approved version.
Any LLM can be replaced with a better one. What cannot be replaced is the GTM knowledge that grounds its answers. Better models still hallucinate your pricing, your process, and your competitive positioning, because that knowledge is not in their training data. Spekit provides the private, governed layer that makes any model accurate for your team.
A knowledge layer is more than a retrieval endpoint. It needs authoring, permissions, approvals, contextual delivery, audit, and gap detection: systems Spekit has spent eight years building and compounding together. A custom build gives you the retrieval layer but not the governance, and the knowledge inside still has to be kept current after launch.
Yes. Through Spekit MCP, reps can generate Deal Rooms and GTM content directly from Claude or ChatGPT without switching to Spekit. Prompts pull from your governed knowledge base, and the output comes back on-brand, with approved messaging, pushed into Spekit and ready to send.
Glean and Claude are powerful horizontal tools, but they retrieve from whatever you connect, including SharePoint folders and Drive docs that carry no governance. Spekit is the GTM-specific knowledge layer with ownership, freshness monitoring, and role-based permissions built in. The two work together through MCP, so those tools retrieve from a governed source instead of raw files.
Spekit MCP works with MCP-compatible hosts and clients in your stack. The tools named on this page are Claude, ChatGPT, Microsoft Copilot, and Glean. Because MCP is an open protocol, Spekit can connect to additional AI tools that support the standard as your AI stack grows.
.svg)
.svg)
