TradeMemory
A persistent memory and audit-trail layer for AI trading agents, recording every trade and its outcome with a tamper-evident hash chain so an agent can recall what happened last time and why.
TradeMemory is a memory layer for AI trading agents, addressing a gap where MCP servers for trading typically handle order execution and price data but not memory: an agent can place a trade but cannot answer what happened the last time it made a similar one. Before trading, an agent can query the memory for what happened last time in a similar market condition; after trading, one call records the outcome across five memory layers, and confidence tracking, drawdown alerts and losing-streak detection can tell the agent when to stop.
Every recorded decision is chained with a tamper-evident SHA-256 hash and RFC 3161 timestamp anchoring, aimed at audit-trail requirements such as MiFID II Article 17 and the EU AI Act's Article 14. It does not execute trades or touch funds itself; it only records and recalls, and works with any market, broker or AI platform through 20 MCP tools. It installs with pip install tradememory-protocol and is in maintenance mode as of August 2026.