Graphs make dynamic agent organizations programmable
Saboo_Shubham_ · x · 2026-08-04
Graphs make agent orgs programmable
The post argues that loops are for how one agent works, while graphs are for how work connects across agents. The key idea is a dynamic agent org: the graph can rewrite itself while the job is still running.
What that looks like in practice
- The author’s setup uses six agents.
- They all read from a shared state layer.
- A coordinator restructures the team based on incoming results.
- The system can promote stronger models, insert extra review when errors pile up, collapse large fan-outs when the budget is too tight, or let a tiny task run solo.
Why it matters
The post’s main claim is that graphs are not just a routing abstraction: they are a way to make the organization of agents itself programmable. The images illustrate a team reshaping itself based on task difficulty, failure signals, and workflow needs.
More from coding & agent
- ARGO open-sources a staged agent loop for traceable AI coding delivery — PossibleOrdinary4572 · 2026-08-04
- Codex can now help fix stubborn Windows driver and startup problems — emollick · 2026-08-04
- AI clone ran a sales front line and surfaced about $3M in opportunities — nikola_mr64990 · 2026-08-04
- A free MCP lets one coding agent DM another after a bug fix — TheThotKiller · 2026-08-04
- AI agents outside coding mostly help with rare beginner-level tasks, not core workflows — chkbd1102 · 2026-08-04
- ScrambleToolBench finds agents still brute-force tools after the map changes — declare-lab · 2026-08-04