MCP Connector

Pilot5 over MCP.

Connect Pilot5 to Claude and run AI council deliberations directly from your conversation.

What is Pilot5?

Pilot5 orchestrates 5 AI personas — The Architect, The Counsel, The Strategist, The Engineer, and The Contrarian — in structured adversarial debate. Each persona runs on a different AI model for genuine diversity of perspective.

Unlike single-model analysis, Pilot5 deliberations include independent analysis, cross-critique, and a synthesis with decision matrix, action plan, confidence scores, and a provenance audit that tracks which claims are evidence-backed vs. inferred.

How to connect

  1. Open Settings → Connectors in Claude.ai
  2. Search for Pilot5 or add a custom connector with URL: https://mcp.pilot5.ai/mcp
  3. Authorize with your Pilot5 account (sign up at pilot5.ai if you don't have one)
  4. Start deliberating from any Claude conversation

Three modes

The Expert (0.1–0.6 credits, ~30 seconds)

Single AI model, fast analysis. Best for quick takes and simple questions.

The A-Team (1.5–2.5 credits, ~5–8 minutes)

5 AI models debate in adversarial rounds with cross-critique. Pauses once after research for your steering input before synthesis. Best for strategic decisions.

The Dream Team (2.5–4.5 credits, ~8–12 minutes)

Same as A-Team but with additional interactive steering rounds. Best for high-stakes decisions where you want to guide the council.

Example prompts

Pricing strategy

“Analyze whether our B2B SaaS should switch from usage-based to seat-based pricing. We have 200 customers, $2M ARR, 40% gross margin, and our biggest competitor just switched to seat-based.”

Fundraising timing

“Our startup is choosing between raising a Series A now at $15M pre-money or waiting 6 months for better metrics. Model the dilution, runway, and strategic positioning for both paths.”

Build vs. buy

“We're deciding between building our ML pipeline in-house vs. using a managed platform. Team of 4 engineers, 18-month runway, need to ship v1 in 3 months.”

Available tools

deliberate — Run a multi-model council deliberation
discover — Refine a vague question before deliberating (free)
check_deliberation_status — Monitor a running deliberation with live progress
get_deliberation_result — Read the full synthesis report
list_deliberations — Search and list past deliberations by topic, mode, or domain
chat — Chat with the Pilot5 assistant for quick questions
search_trusted_sources — Search verified institutional sources (Eurostat, OECD, DGFiP, etc.)
search_memories — Search insights from past deliberations
critique_deliberation — Grade and critique a completed deliberation

24 tools total. Run check_credits to see the full list. The deliberation report is also exposed as an MCP resource at mcp://pilot5.ai/deliberation/{id} so Claude can attach it as a document artifact.

Support

Questions or issues? Contact contact@pilot5.ai. For privacy requests: privacy@pilot5.ai. Enterprise & DPAs: enterprise@pilot5.ai.

See also the Privacy Policy, Terms of Service, and Data Processing Agreement.

Once the connector is authorised, you can run a deliberation from any Claude conversation — or test it on the web first to see what the output looks like end-to-end.