Open‑source finance agent pack
‘Awesome Finance Skills’ — an open‑source pack — is being shared to turn AI agents into pro analysts: it pulls real‑time news, forecasts prices via a ‘Kronos’ model, and auto‑generates reports; it’s MIT‑licensed and designed to work with Claude and Gemini. If you build market‑monitoring agents, this is a ready‑made kit to accelerate forecasting workflows. (x.com)
The package is published on GitHub under RKiding/Awesome‑finance‑skills and the repository shows roughly 330 stars, 27 forks and a README update listed three weeks ago. (github.com)) The project’s top‑level metadata and several included skill license files identify the codebase as Apache‑2.0 (not MIT). (github.com)) The repo bundles named skills such as alphaear‑news, alphaear‑predictor (Kronos‑based forecasting), alphaear‑reporter and alphaear‑logic‑visualizer, and its README lists source feeds including Cailian, The Wall Street Journal, Weibo and Polymarket. (github.com)) Installation options include a one‑step npx installer and manual cloning, and the README documents integration paths for multiple agent frameworks with explicit directories for Gemini/Antigravity and Claude (for example ~/.gemini/antigravity/global_skills/ and ~/.claude/skills/). (github.com)) The Kronos time‑series model used by the predictor is an open‑source foundation model pre‑trained on a corpus of over 12 billion K‑line records from 45 global exchanges and is released under an MIT license with model checkpoints hosted on GitHub and Hugging Face. (arxiv.org)) Kronos’s repo shows broad community uptake (about 11.3k stars and 2.4k forks) and a live demo for forecasts, while the RKiding skill pack is maintained alongside related projects such as DeepEar and publishes release assets and example SKILL.md files. (github.com))
