Couches architecturales
Cinq préoccupations orthogonales qui se combinent en un agent complet
L1 Tools & Execution
Ce que l'agent PEUT faire. La base : les outils donnent au modèle des capacités pour interagir avec le monde.
L2 Planning & Coordination
Comment le travail est organisé. Des listes de tâches simples aux tableaux de dépendances partagés entre agents.
TodoWrite
Plan Before You Act
An agent without a plan drifts; list the steps first, then execute
Subagents
Clean Context Per Subtask
Subagents use independent messages[], keeping the main conversation clean
Skills
Load on Demand
Inject knowledge via tool_result when needed, not upfront in the system prompt
Tasks
Task Graph + Dependencies
A file-based task graph with ordering, parallelism, and dependencies -- the coordination backbone for multi-agent work
L3 Memory Management
Garder le contexte dans les limites. Stratégies de compression qui permettent aux agents de travailler infiniment sans perdre de cohérence.
L4 Concurrency
Exécution non bloquante. Threads d'arrière-plan et bus de notifications pour le travail parallèle.
L5 Collaboration
Coordination multi-agents. Équipes, messagerie et co-agents autonomes qui pensent par eux-mêmes.
Agent Teams
Teammates + Mailboxes
When one agent can't finish, delegate to persistent teammates via async mailboxes
Team Protocols
Shared Communication Rules
One request-response pattern drives all team negotiation
Autonomous Agents
Scan Board, Claim Tasks
Teammates scan the board and claim tasks themselves; no need for the lead to assign each one
Worktree + Task Isolation
Isolate by Directory
Each works in its own directory; tasks manage goals, worktrees manage directories, bound by ID