jwbron · jwbron · May 21, 2026 · May 20, 2026 · May 20, 2026 · May 20, 2026
diff --git a/.egg-state/agent-outputs/2623-architect-output.json b/.egg-state/agent-outputs/2623-architect-output.json
diff --git a/.egg-state/agent-outputs/2623-risk_analyst-output.json b/.egg-state/agent-outputs/2623-risk_analyst-output.json
diff --git a/.egg-state/agent-outputs/2717-architect-output.json b/.egg-state/agent-outputs/2717-architect-output.json
diff --git a/.egg-state/agent-outputs/2717-risk_analyst-output.json b/.egg-state/agent-outputs/2717-risk_analyst-output.json
diff --git a/.egg-state/brc-history/2623-plan.json b/.egg-state/brc-history/2623-plan.json
diff --git a/.egg-state/brc-history/2623-plan.md b/.egg-state/brc-history/2623-plan.md
diff --git a/.egg-state/brc-history/2623-refine.json b/.egg-state/brc-history/2623-refine.json
diff --git a/.egg-state/brc-history/2623-refine.md b/.egg-state/brc-history/2623-refine.md
diff --git a/.egg-state/brc-history/2717-implement-slice-1.json b/.egg-state/brc-history/2717-implement-slice-1.json
diff --git a/.egg-state/brc-history/2717-implement-slice-1.md b/.egg-state/brc-history/2717-implement-slice-1.md
diff --git a/.egg-state/brc-history/2717-implement-slice-2.json b/.egg-state/brc-history/2717-implement-slice-2.json
diff --git a/.egg-state/brc-history/2717-implement-slice-2.md b/.egg-state/brc-history/2717-implement-slice-2.md
diff --git a/.egg-state/brc-history/2717-plan.json b/.egg-state/brc-history/2717-plan.json
diff --git a/.egg-state/brc-history/2717-plan.md b/.egg-state/brc-history/2717-plan.md
diff --git a/.egg-state/brc-history/2717-refine.json b/.egg-state/brc-history/2717-refine.json
diff --git a/.egg-state/brc-history/2717-refine.md b/.egg-state/brc-history/2717-refine.md
diff --git a/.egg-state/contracts/issue-2623.json b/.egg-state/contracts/issue-2623.json
diff --git a/.egg-state/contracts/issue-2717.json b/.egg-state/contracts/issue-2717.json
diff --git a/.egg-state/drafts/2623-analysis.md b/.egg-state/drafts/2623-analysis.md
diff --git a/.egg-state/drafts/2623-plan.md b/.egg-state/drafts/2623-plan.md
diff --git a/.egg-state/drafts/2717-analysis.md b/.egg-state/drafts/2717-analysis.md
diff --git a/.egg-state/drafts/2717-plan.md b/.egg-state/drafts/2717-plan.md
diff --git a/docs/architecture/README.md b/docs/architecture/README.md
@@ -191,7 +191,6 @@ The SDLC pipeline orchestrates agent-based development with structurally enforce
 - [Credential Injection](credential-injection.md) - Zero-credential sandbox with API key proxy
 - [Network Isolation](network-isolation.md) - Public/private network modes
 - [SDLC Pipeline](sdlc-pipeline.md) - Structurally enforced agent checkpoints
-- [Claude Code Substrate](claude-code-substrate.md) - Substrate-swap walking skeleton (#2623): four `Protocol`s (`AgentSpawner` / `MessageBus` / `PolicyEnforcer` / `WorktreeManager`) under `orchestrator/substrate/`, env-var-selected via `EGG_SUBSTRATE`, with a working `K3sSpawnerAdapter` shim and Claude-Code-native implementations (`ClaudeCodeSpawner`, `InProcessMessageBus`, `PreToolUseHookPolicy`, `LocalWorktreeManager`); ships an in-process orchestrator generator (`run_pipeline_in_process`) yielding `HITLDecision` objects (cq-7 heredoc-HITL). The ADR records the trust-context shift (R1), PreToolUse-hook fallback to MCP-validator-side enforcement (R2), subagent context-budget regression vs `max_turns:1000` (R7), and the `EGG_PIPELINE_MAX_AGENT_INVOCATIONS` cost-cap recommendation (REC5); the four interfaces are marked `# v0.x — unstable until ≥3 roles exercise`; a "Follow-up issue draft" appendix captures the deferred rollout (plan / implement / pr phases, full 5-issue conformance matrix, perf budget)
 - [Declarative Setup](declarative-setup.md) - Python-based setup
 - [Logging](logging.md) - Structured JSON logging
 - [Integration-Test Trust Boundary](integration-test-trust-boundary.md) - Test execution contexts (in-sandbox-agent / trusted-CI-runner / human-operator), fixture tier table, and hard-NACK rules for plan-phase Primitive-Existence (§9) and Trust-Boundary (§10) audits

diff --git a/docs/architecture/claude-code-substrate.md b/docs/architecture/claude-code-substrate.md
diff --git a/docs/development/STRUCTURE.md b/docs/development/STRUCTURE.md
@@ -139,21 +139,6 @@ orchestrator/
 ├── status_reporter.py      # Real-time status reporter for collaborators
 ├── unified_sse.py          # Unified SSE stream for all pipelines
 ├── webhooks.py             # GitHub webhook handlers
-├── substrate/              # Substrate-swap abstraction layer (walking-skeleton #2623): four Protocol interfaces + env-var-selected factory
-│   ├── __init__.py         # `select_substrate(env)` factory; reads `EGG_SUBSTRATE` (`k3s` or `claude-code`)
-│   ├── spawner.py          # `AgentSpawner` protocol + `AgentResult` dataclass
-│   ├── message_bus.py      # `MessageBus` protocol
-│   ├── policy.py           # `PolicyEnforcer` protocol
-│   ├── worktree.py         # `WorktreeManager` protocol
-│   ├── k3s_adapter.py      # `K3sSpawnerAdapter` shim wrapping `KubernetesSpawner.create_concurrent_spawn_fn`
-│   ├── in_process.py       # `run_pipeline_in_process()` generator entry point (claude-code substrate only)
-│   └── claude_code/        # Claude-Code-native implementations
-│       ├── spawner.py      # `ClaudeCodeSpawner` — runs `egg_harness.run_agent` in-process (Agent-tool dispatch is a follow-up; see ADR "Open work")
-│       ├── hook_entry.py   # Standalone `python3 -m` PreToolUse hook script invoked by Claude Code (Bash/Write/Edit parser; fail-closed on ambiguous shapes)
-│       ├── message_bus.py  # `InProcessMessageBus` — subclasses `MessageStore` (in-memory)
-│       ├── policy.py       # `PolicyEnforcer` adapter (`check_write` + `install`) wrapping the `hook_entry.py` script
-│       ├── settings.template.json # Claude Code settings template registering `hook_entry.py` as the PreToolUse hook
-│       └── worktree.py     # `LocalWorktreeManager` — per-agent worktrees under `~/.egg-worktrees/`
 ├── overseer/               # Overseer agent package (LLM-powered tier of pipeline health monitoring)
 │   ├── classifier.py       # Haiku-tier classifiers (stall, loop, error triage, off-track detection)
 │   ├── decision_maker.py   # Sonnet/Opus-tier decision-maker (corrective actions, redirect messages)

diff --git a/docs/index.md b/docs/index.md
@@ -31,7 +31,6 @@ This index helps both humans and LLMs navigate the documentation efficiently.
 | [The Agentic Feedback Loop](architecture/agentic-feedback-loop.md) | The foundational work-review-feedback cycle that drives quality |
 | [Why egg Works](architecture/collaboration-effectiveness.md) | How the public, sandboxed, async model delivers safety and collaboration |
 | [Integration-Test Trust Boundary](architecture/integration-test-trust-boundary.md) | Test execution contexts (in-sandbox-agent / trusted-CI-runner / human-operator) and fixture tiers; authoritative reference for plan-phase Trust-Boundary Audit (#2594) |
-| [Claude Code Substrate](architecture/claude-code-substrate.md) | Substrate-swap walking-skeleton (#2623): four `Protocol`s (`AgentSpawner` / `MessageBus` / `PolicyEnforcer` / `WorktreeManager`) under `orchestrator/substrate/`, env-var-selected via `EGG_SUBSTRATE`, with Claude-Code-native implementations and an in-process orchestrator generator (`run_pipeline_in_process`) |
 
 ### Development