Funes turned coding-agent traces into a searchable memory layer
The project addressed the gap between keeping old sessions and retrieving useful context from them.
Article·4 minEXPLORE THE ECOSYSTEM
Agents
Useful tools, memory, and better workflows for AI agents.
8 articlesNewest announcements first
Article·4 min
Agent Reproductions Need Better Questions, Not Just More Runs
A large ICML 2026 reproduction challenge shows where coding agents can widen research scrutiny—and why claim selection, independent runs, scale checks, and human judgment still determine whether the evidence is trustworthy.
Article·5 min
The hf CLI was redesigned with coding agents in mind
A more discoverable command-line interface aimed to reduce the work needed for automated Hub tasks.
Article·4 min
Training an agent to reason inside a notebook
Jupyter Agent connects code execution, training data, and evaluation for analytical tasks.
Article·4 min
A small agent loop makes MCP easier to understand
Tiny Agents shows the basic connection between an inference client and a tool server.
Article·4 min
What an open research agent adds around a language model
An early open research-agent effort makes the surrounding tools and execution loop visible.
Article·4 min
Smolagents learns to work with what a page looks like
Vision inputs let an agent use layout and visual feedback, not just extracted text.
Article·4 min
smolagents kept the first agent loop deliberately small
A lightweight library made tool-using agents easier to inspect and experiment with.
Article·4 min