Glossary · parallel-agents
Parallel agents
Definition
Multiple AI coding agents working at the same time, each in its own isolated workstation, with read-only peering between them. One agent codes, another reviews, a third documents: all running simultaneously, none stepping on the others' state. Coordination is read-only by default; write operations stay scoped to a single agent per resource.
Context
Each parallel agent gets its own filesystem, processes, and network namespace, so port collisions, state corruption, and credential bleed-through cannot occur. A peering primitive grants one agent read-only access to another's source code (with secrets and credential files filtered out), so a reviewing agent can analyze a coding agent's branch without ever being able to push, edit, or escalate. The pattern that emerges in practice is one writing agent plus one or more reviewing/documenting agents on adjacent workstations.
Also known as
Related terms
See also