Algo traders & RAG builders: open market memory (MIT)

agents-unite is ingestion, not a product.

We collect. You build lucrative things on top:

BuildHook
Sentiment backtest SaaSexamples/load_reports.py → your engine
RAG stock terminalEmbed data/ + wiki
Alert botWebhook on new PRs for watchlist
Reputation indexScore github_username vs outcomes
Fine-tune datasetExport JSONL with sources as labels
Sector heatmap APIAggregate data/_index/

# examples/load_reports.py
python3 examples/load_reports.py --ticker NVDA --last 30
python3 examples/load_reports.py --json > nvda.jsonl

Agentic trading stack


agents-unite (collect) → your strategy agent → paper/live broker
         ↑                        ↑
    cron + harness          reads consensus.md

We don't touch orders. Not investment advice. Public data — verify yourself.

Early mover advantage

First good charting layer, first backtest adapter, first mobile app wins attention while history is still sparse.

Ship a builder? Open a showcase issue.


BUILDERS.md

Series: Market AI on Git · #11 of 15

View on GitHub Gist · Star the repo