Investor Diligence
Click on "Deploy Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type
@followed by the MCP server name and your instructions, e.g., "@Investor Diligencerun the investor council on my pitch deck and summarize the risks"
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
Investor diligence — Convex deployment
Analyze pitch decks and transcripts through a Convex-backed website and authenticated HTTP MCP service. Convex owns source files, versioned evidence, analysis state, and durable workflow orchestration. Backend actions call xAI directly; the website and MCP adapter share case authorization.
Use Node 22.12+ (below 25) and pnpm 10.30.3:
pnpm install --frozen-lockfile
pnpm setup:check
pnpm setup:convex # push to the configured development deployment
pnpm dev # website and HTTP MCP adapterRun pnpm dev:convex separately to watch and deploy backend edits. Configure public VITE_CONVEX_URL in the web app's environment; Vite does not load root .env.local. See setup, browser authentication, and web integration. pnpm dev:hosted is an alias for the same website/MCP development path.
The council runs intake → context brief → ten parallel specialist lenses → Devil's Advocate → ranked report. Only intake receives files. Outputs retain explicit source references and undergo schema and evidence checks. PDF quotations and page accuracy still require evaluation. Convex stores attempts, usage, immutable source/configuration snapshots, and intermediate artifacts; runs continue independently of the client connection.
Live analysis requires XAI_API_KEY and LIVE_ANALYSIS_ENABLED=true on Convex. Tests simulate xAI; paid inference, evidence quality, and named-client compatibility remain unverified. Hosted MCP fails closed until OAuth and the identity bridge are configured; WorkOS account setup and browser callback UI remain outstanding. Browser password recovery and chat are also unfinished.
render.yaml describes website/MCP hosting separately from the Convex deployment. Restoring Convex does not itself publish those services. See MCP specification and workflow details.
The disk-backed runtime remains available through pnpm dev:local, pnpm setup:check:local, and pnpm start:local; see local setup. Existing stdio client configurations still use that local runtime. Local runs are not migrated into Convex automatically.
pnpm check # configuration, types, tests, builds, formattingSee plan. The earlier hosted implementation record preserves detailed milestones and historical verification results.
This server cannot be deployed
Maintenance
Related MCP Connectors
MCP server for VC pitch-deck scoring, thesis-fit matching, and deal-flow management.
Pre-diligence AI for founders, investors, and firms — multi-agent pitch analysis and deal flow.
Tenant-scoped evidence intake and controlled AI context over MCP.
Venture intelligence for autonomous agents with discovery, scoring, and workspace automation.
Related MCP Servers
- FlicenseAqualityAmaintenanceGovernance/control plane for MCP-enabled coding-agent workflows with validation, findings, approvals, budgets, and proof bundles.5511-
- AlicenseAqualityAmaintenanceProvides Web3 founders with tools for project scoring, portfolio matching, fund matching, hackathon discovery, jurisdiction analysis, Twitter audit, and smart contract security review. Works with MCP-compatible assistants without transmitting file contents.8100 npm7MIT
- AlicenseAqualityDmaintenanceLocal-first MCP server that simulates an investment committee on your pitch materials to assess funding readiness, integrating with Obra CTO for technical evidence.332 npmApache 2.0
- AlicenseNot gradedqualityBmaintenanceAn MCP-native agentic platform orchestrating planner/executor/critic agents over hybrid RAG with three-tier memory, budget enforcement, safety guardrails, and full observability. It exposes all capabilities as MCP tools, enabling natural-language control of document ingestion, retrieval-augmented generation, and multi-step AI workflows.MIT