Day page
19 May 2026
Paper summary
A survey that treats code as the infrastructure of agents, not just their output: the interface agents act through, the mechanism for planning, memory and verification, and the shared artefact that coordinates multiple agents.
- In agent systems, code has become the medium for reasoning, acting, modelling the environment and verifying by execution.
- The survey is organised in three layers: harness interface, harness mechanisms, and scaling to multi-agent settings.
Read the post →
Study
LangGraph models an application as a state machine: typed state, nodes that update it, and conditional edges, used here to build a small corrective-retrieval loop that rewrites the query when retrieval looks weak.
- A LangGraph app is a graph: nodes are functions that update a shared state, and edges (including conditional ones) decide what runs next.
- Loops are allowed, which is what agent-style "try, check, retry" workflows need.
Read the post →
AI updates
Google DeepMind introduced Gemini 3.5 with 3.5 Flash first, reporting it beats Gemini 3.1 Pro on agentic and coding benchmarks (Terminal-Bench 2.1 76.2%, MCP Atlas 83.6%) at about 4× the output speed of other frontier models.
- A Flash model reported to beat the previous Pro on coding and agent benchmarks.
- MCP Atlas, an agent benchmark built around the Model Context Protocol, is among the reported results.
Read the post →