Official docs distinguish one-way subagent delegation from experimental peer-capable Agent Teams and state that teammates do not receive automatic worktree isolation.
claude-code-2026-08-10-docs →MECHANISM · ORCHESTRATION · EMERGING
Subagent Orchestration
Delegate bounded work into isolated contexts, control capabilities and lifecycle, and merge results without confusing ownership or workspace state.
1 lessons6 agent snapshots0 experiments
orchestrationsubagent-orchestrationproblem → policy → evidence
L0Intuitionpartial
L1Buildmissing
L2Engineeringpartial
L3Architecturepartial
L4Researchpartial
DEPENDENCY GRAPH · READER PATH
What feeds it, and what does it unlock?
Arrows show learning and design dependencies, not runtime data flow. Follow any node to continue through the Atlas.
Prerequisites
subagent-orchestrationSubagent OrchestrationCurrent research boundary
UnlocksNo dependents yet
Delegate bounded work into isolated contexts, control capabilities and lifecycle, and merge results without confusing ownership or workspace state.
The structured record is in the Atlas, but the L0–L4 article has not passed content review.
AGENT MAPPING · EVIDENCE ONLY
Snapshot implementations
An implementation enters the map only when both a snapshot and claims exist; unknowns remain visible.
1 evidence-backed claim.
codex-2026-08-10-source →1 evidence-backed claim.
grok-build-2026-08-10-source →1 evidence-backed claim.
opencode-2026-08-10-source →The source-backed result is an intentional absence from core plus documented extension points.
pi-2026-08-10-source →1 evidence-backed claim.
reasonix-2026-08-10-source →No implementation currently clears the Snapshot + Claim threshold; this is not a claim that the product lacks the capability.
Inspect coverage boundary →CLAIMS · EVIDENCE LEDGER7 RECORDS
official-docreviewed
claude-code-doc-agent-teams
Claude Code's official Agent Teams documentation describes experimental lead-and-teammate coordination with independent context windows, a shared task list, direct messaging, and no automatic worktree isolation.
- Agent
- Claude Code
- Snapshot
- claude-code-2026-08-10-docs
- Reviewed
- 2026-08-10
official-docreviewed
claude-code-doc-subagents
Claude Code's official documentation describes subagents with separate context windows, configurable prompts, tools, models, permissions, hooks, skills, depth and turn limits, and optional temporary Git worktree isolation.
- Agent
- Claude Code
- Snapshot
- claude-code-2026-08-10-docs
- Reviewed
- 2026-08-10
source + sourcereviewed
codex-source-subagent-routing
The pinned Codex core registers multi-agent tools for spawning and communicating with child agents and implements spawn through a dedicated agent-control boundary rather than inside the main sampling loop.
- Agent
- Codex
- Snapshot
- codex-2026-08-10-source
- Reviewed
- 2026-08-10
source + sourcereviewed
grok-build-source-subagent-runtime
Grok Build's pinned source resolves named subagents and resume identity, while task execution uses separate child session identity and explicit isolation or worktree metadata rather than an inline prompt macro.
- Agent
- Grok Build
- Snapshot
- grok-build-2026-08-10-source
- Reviewed
- 2026-08-10
sourcereviewed
opencode-source-subagent-sessions
OpenCode's pinned task tool limits nesting, derives child permissions, creates parent-linked child sessions, and supports foreground or background tasks with explicit metadata.
- Agent
- OpenCode
- Snapshot
- opencode-2026-08-10-source
- Reviewed
- 2026-08-10
sourcereviewed
pi-source-orchestration-boundary
Pi intentionally omits a built-in subagent orchestration policy; its official project guidance points users to separate Pi processes or extensions so delegation semantics remain outside the minimal core.
- Agent
- Pi
- Snapshot
- pi-2026-08-10-source
- Reviewed
- 2026-08-10
sourcereviewed
reasonix-source-subagents
Reasonix constructs a filtered child tool registry, supports a read-only child profile, caps recursive delegation by depth, hides parent-only orchestration and job tools, and gives each child a private temporary-session boundary.
- Agent
- Reasonix
- Snapshot
- reasonix-2026-08-10-source
- Reviewed
- 2026-08-10
EXPERIMENTS · EXERCISES
What has actually been tested?
Formal experiments are separate from course exercises. Exercises can validate the reference implementation but cannot replace Native evidence from a real agent.
Formal experiment · not registered
No experiment record means untested, not passed. The course checks below cover only the deterministic reference harness.
s01-agent-loopobserve
Verify the Golden TraceThe committed JSONL exactly matches a real demo run and passes the trace schema.
python3 -m curriculum.golden verify s01-agent-loops01-agent-loopmodify
Preserve loop invariantsA completed run keeps contiguous sequence numbers and ends with session.stop.
python3 -m unittest curriculum.tests.test_vertical_slice.VerticalSliceTests.test_s01_minimal_loop -v01
Which context, permissions, and instructions should a child inherit?
02
When is a worktree a sufficient isolation boundary?
03
How should cancellation, partial results, and nested delegation be represented in traces?