We just open-sourced Octagon VC Agents.
It’s an MCP server that lets you run AI-driven simulations of VCs like Fred Wilson, Peter Thiel, Marc Andreessen, and Reid Hoffman — designed for real research tasks like pitch feedback, diligence drills, and term sheet practice.
It’s powered by OpenAI’s new Agents SDK, so you get true agentic workflows out of the box:
Multi-agent reasoning
Context handoff
Tool use orchestration
Easy persona customization via markdown
Each agent is continuously enriched with real private markets data from Octagon (startup deals, valuations, founder profiles), so the answers stay grounded — not just hallucinated LLM output.
Use cases:
Side-by-side deal critiques from different VC perspectives
Reality-check startup metrics before fundraising
Simulate partner meeting debates or warm intro strategies
Create new investor personas tuned to your fund’s style
Agents can be extended easily — just modify their investment philosophy, cognitive biases, or communication styles in the persona files. You can also plug in your own orchestration logic if you want more complex decision chains.
You can also checkout a hosted custom gpt version here:
https://chatgpt.com/g/g-680c1eddd1448191bb4ed7e09485270f-vc-...
Would love feedback, ideas, or if anyone wants to collaborate on expanding multi-agent debate flows or new use cases!