Before a task starts, the system pre-assembles everything the agent needs:
- The task plan (goal, steps, must-haves)
- Compressed summaries from dependency slices
- Milestone-level context and decisions
- Continue-here data if resuming interrupted work
This is injected automatically. The agent never has to
grep for project structure, read state files to figure out where it is, or search for what was built in prior slices. If it does, the context assembly is broken — that's a bug, not a workflow.
The goal is zero discovery calls. Every token the agent spends on "where am I, what exists, what was decided" is a token not spent on actual implementation.
......
This would let the agent stay in the context window zone -- the smart zone more easily. But also this feels exactly like Claude Code's planning, then context clearing, then implementing. What's the difference? determinism where possible... maybe?