Letting an agent deploy trading strategies over MCP, then grading it with something it cannot touch
QuanTradin · reddit · 2026-09-24
The author exposes 22 tools over MCP so an agent (tested with Claude Code and Claude Desktop via mcp-remote) can build, backtest and deploy trading strategies to a paper account autonomously. Key insight: agents are good at generating candidate strategies but terrible at judging their own output. The fix: a separate grader the agent can read but cannot influence. Of 70 strategies, 1 has a LIVE verdict, 1 SHADOW, 1 SHELF. Three reusable mechanisms: per-account key pairs, capability scopes (read/backtest/deploy) bound to each pair, and an owner-only freeze the agent can see but only a human can set. All paper trading, no real money.
More from coding & agent
- Ex-Google engineer built a boss fight and spaceflight into his personal site — kieranklaassen · 2026-09-24
- Dan Grover: agent memory systems write and retrieve memories sparingly and arbitrarily — DanGrover · 2026-09-24
- Continuous Benchmarks: Treat Evals Like Software, Not Static Artifacts — kenbwork · 2026-09-24
- Measured: Claude Code silently skips AGENTS.md when telemetry is off — steipete · 2026-09-24
- GPT-6 self-critiques its design with 28 fixes, rebuilds a landing page in 11 minutes — PrajwalTomar_ · 2026-09-24
- After Interviewing 50 Developers, the Top 8 Mistakes in AI-Assisted Coding — dfinke · 2026-09-24