we open-sourced Parasol's trading engine — here's why

the short version

Parasol's full codebase is now MIT-licensed and public on GitHub. the trading engine, the risk management system, the scanner, the scoring algorithms, the MCP server, the SDK — all of it. anyone can read it, fork it, build on it.

we did this with grant funding from the Solana Foundation via Superteam UK.

why open source a trading engine

the obvious question: if your trading strategies are public, what stops someone from copying them?

nothing. they can.

but the strategies aren't the moat. the moat is the accumulated learning data (which lives in our database, not in Git), the speed of iteration (we built this — we can tune it faster than someone reading cold code), and the operational knowledge of what actually works on Solana memecoins at 3am UTC when liquidity dries up and the spreads blow out.

more importantly, we took grant money from the Solana Foundation to build infrastructure for the public good. keeping that behind a proprietary license would be dishonest. the grant exists so that builders on Solana have better tools. MIT is the clearest way to deliver on that.

what the grant covered

Superteam UK awarded us $7,500 USDG under their Startup — Early Scaling category. the deliverables:

all of these are complete and shipped to production on parasol.so.

what's actually in the repo

this isn't a marketing SDK wrapper with the real logic hidden on a server. the full production codebase is public:

10 MCP tools that any AI agent (Claude, GPT, Cursor, custom) can call via the Model Context Protocol:

8 REST API endpoints for programmatic access: key management, agent CRUD, trade execution, signal injection, scanner results, performance analytics.

7 trading strategies — momentum, mean reversion, breakout, volume climax, whale wake, renewed momentum, and pressure zone — each with explicit scoring logic, gate conditions, and entry/exit rules.

6-layer manipulation filter — checks holder concentration, sniping patterns, liquidity depth, wash trading signals, bundle detection, and deployer history before any token reaches the scoring engine.

adaptive risk management — daily and weekly loss limits, peak drawdown kill switch, portfolio heat tracking, concept drift detection that reduces position sizes when market conditions shift against the strategies, cash reserve requirements, consecutive loss cooldowns, and winner re-entry blocks.

adaptive learning — feature snapshots captured at every trade entry, paired with outcomes on close. after enough samples, the learner adjusts strategy weights: boosting signals correlated with winners, suppressing those correlated with losers. operates at three levels: global, regime-specific, and regime+strategy-specific.

3 free trading calculators at parasol.so/tools — position sizing, funding rate, and liquidation price — that work without any signup and drive organic search traffic.

what this means for builders

if you're building on Solana and need to:

you can. the code is MIT. take what you need.

the Phantom integration

one thing we're particularly proud of: Parasol works as a safety layer for Phantom wallet users via MCP.

the flow: an AI agent (Claude, Cursor, any MCP client) calls Parasol's parasol_safe_trade_check before executing a swap through Phantom's MCP server. Parasol checks rug risk, honeypot status, holder quality, and deployer history. if the token is dangerous, the agent stops before the trade reaches the wallet. if it's safe, Phantom previews the swap and the user confirms.

this is what "public good" looks like in practice. Phantom has millions of users. if even a fraction of them avoid one rug pull because an AI agent checked with Parasol first, that's real money saved for real people.

the numbers

what's next

the Colosseum Frontier hackathon submission is in progress. the SDK is live. the documentation is published. if you want to build on Parasol, start here:

if you're using Parasol in your project, we'd love to hear about it. reach out on X @ParasolSolana.