Ibitlabs Mcp
Let your agent read live balances, trades, shadow rules, and saga receipts from the public iBitLabs $1k→$10k AI trading experiment.
Overview
iBitLabs MCP is a MCP server for the Grow phase that exposes live trading experiment balances, trades, shadow rules, and saga receipts to your agent.
What is this MCP server?
- stdio npm package ibitlabs-mcp at version 0.2.2 for local MCP wiring
- Live receipts: real balance and trade history from an ongoing AI experiment
- Shadow rules and saga context for comparing agent decisions to documented policy
- Public experiment framing: $1k to $10k growth narrative for benchmarking agents
- GitHub source under AgentBonnybb/ibitlabs mcp-server subfolder
- npm package version 0.2.2 with stdio transport
- Experiment narrative anchor: $1k→$10k AI trading journey
- Repository: github.com/AgentBonnybb/ibitlabs (mcp-server subfolder)
What problem does it solve?
Finance and trading agents hallucinate performance unless you pipe in authoritative, timestamped account and trade receipts.
Who is it for?
Solo builders running or documenting AI trading experiments who want MCP-accessible live receipts for analysis, demos, or content.
Skip if: Production order routing, licensed advisory workflows, or teams that need a regulated execution stack instead of public experiment telemetry.
What do I get? / Deliverables
Your agent can cite real experiment state—balance, trades, rules, and saga—from iBitLabs when analyzing or narrating the $1k→$10k run.
- Agent-queryable live balance and trade history from the experiment
- Shadow-rule and saga context for decision postmortems
- Grounded narratives for grow-phase analytics and content
Recommended MCP Servers
Journey fit
After you ship an autonomous trading or finance agent, you need grounded metrics and audit trails to learn what worked—this MCP surfaces live experiment data for that loop. Analytics is the right shelf because the server exposes balances, trades, and narrative saga as observability inputs, not build-time scaffolding.
How it compares
Read-only finance telemetry MCP, not a loan marketplace or backtesting CLI skill.
Common Questions / FAQ
Who is iBitLabs MCP for?
Indie builders and agent authors who follow the iBitLabs AI trading experiment and want Claude or Cursor to read live balances, trades, and saga context.
When should I use iBitLabs MCP?
Use it in Grow when you are analyzing experiment performance, writing transparent updates, or tuning agent rules against real trade history.
How do I add iBitLabs MCP to my agent?
Install the npm package ibitlabs-mcp (0.2.2), configure stdio transport in your MCP settings, and run the server locally per the ibitlabs repository mcp-server folder.