Skip to content

Upsonic

AI Agents & Assistants

A Python framework for building autonomous AI agents

Python Latest v0.77.3 · 15d ago Security brief →

Features

  • Build autonomous AI agents with simple Python APIs
  • Support for prebuilt community agents and custom tool integration
  • Unified OCR interface with multiple engine back‑ends

Recent releases

View all 31 releases →
No immediate action
v0.77.3 Bug fix

Centralized usage registry

No immediate action
v0.77.2 Bug fix

Chat & messages bugfix

Config change
v0.77.0 Bug fix
Auth

Sentry opt‑in

0.76.2 New feature
Notable features
  • AppliedScientist accepts any research_source — local file, URL, git/Kaggle/arXiv/Hugging Face link, or plain text idea
  • scientist.new_experiment() auto-derives inputs from research_source, current_notebook, or current_data by detecting local paths
  • experiments_directory is now optional with default value ./experiments
Full changelog

Improvements:

  • AppliedScientist now accepts any research_source — local file, URL, git / Kaggle / arXiv / Hugging Face link, or a plain-text idea. No hardcoded scheme detection on the Python side; the agent skill handles fetching with whichever tool fits (cp / git clone / curl / kaggle CLI / huggingface-cli / …). For text ideas, the description is saved verbatim and Phase 2 turns it into a concrete method plan.
  • scientist.new_experiment(...) auto-derives inputs from research_source / current_notebook / current_data — any value that points at an existing local path is copied into the workspace.
  • experiments_directory is now optional and defaults to ./experiments.

Full Changelog: https://github.com/Upsonic/Upsonic/compare/v0.76.1...v0.76.2

Pull Requests:

v0.76.1 Breaking risk
Breaking changes
  • AppliedScientist: research_paper parameter renamed to research_source
Notable features
  • research_source parameter accepts PDF, Markdown, HTML, .ipynb, text files, web URLs, git repositories, and Kaggle pages
  • Docs repository added as git submodule
Full changelog

Improvements:

  • AppliedScientist research_paperresearch_source: The prebuilt Applied Scientist agent now accepts any reference describing a new method — local file (PDF, Markdown, HTML, .ipynb, text), web URL (blog post, arXiv, documentation), git repository URL, or Kaggle notebook / dataset page — not just a PDF. The agent detects the source kind at Phase 0 and materializes it into the experiment folder before reading it in Phase 2.
  • Upsonic/Docs submodule: Added the public docs repository as a git submodule at Docs/ so the docs site can be kept in lockstep with code changes.

Full Changelog: https://github.com/Upsonic/Upsonic/compare/v0.76.0...v0.76.1

Pull Requests:

Weekly OSS security release digest.

The CVE patches and breaking changes that affected production tools this week. One email, every Sunday.

No spam, unsubscribe anytime.

About

Stars
7,868
Forks
731
Languages
Python Jupyter Notebook Makefile

Install & Platforms

Install via
pip

Community & Support

Beta — feedback welcome: [email protected]