diff --git a/checkpoint.json b/checkpoint.json
new file mode 100644
index 000000000..d9bccd860
--- /dev/null
+++ b/checkpoint.json
@@ -0,0 +1,52 @@
+{
+  "timestamp": "2026-03-16T01:58:37.664339Z",
+  "current_node": "toolchain",
+  "completed_nodes": [
+    "start",
+    "toolchain"
+  ],
+  "node_retries": {
+    "start": 1,
+    "toolchain": 1
+  },
+  "context_values": {
+    "current.preamble": "Goal: # Wire up missing hook invocations\n\n## Context\n\nFive `HookEvent` variants exist in the enum and are documented in `docs/agents/hooks.mdx`, but `run_hooks()` is never called for them in the engine. Users can configure hooks for these events, but they silently never fire.\n\n**Events to wire up:** `StageRetrying`, `ParallelStart`, `ParallelComplete`\n**Events to mark as reserved:** `SandboxReady`, `SandboxCleanup` (sandbox lifecycle is managed outside the engine; wiring these requires significant architecture changes)\n\n## Changes\n\n### 1. Add StageRetrying hook calls in `engine.rs`\n\nFile: `lib/crates/fabro-workflows/src/engine.rs`\n\nTwo sites in `execute_with_retry`, both immediately after `WorkflowRunEvent::StageRetrying` emission and before `tokio::time::sleep(delay).await`:\n\n- **Site A (~line 1127):** error-retry path\n- **Site B (~line 1155):** explicit Retry status path\n\nPattern (same for both sites):\n```rust\n{\n    let mut hook_ctx = HookContext::new(\n        HookEvent::StageRetrying,\n        context.run_id(),\n        graph.name.clone(),\n    );\n    hook_ctx.node_id = Some(node.id.clone());\n    hook_ctx.node_label = Some(node.label().to_string());\n    hook_ctx.handler_type = node.handler_type().map(String::from);\n    hook_ctx.attempt = Some(usize::try_from(attempt).unwrap_or(usize::MAX));\n    hook_ctx.max_attempts = Some(\n        usize::try_from(policy.max_attempts).unwrap_or(usize::MAX),\n    );\n    let _ = self.run_hooks(&hook_ctx, None).await;\n}\n```\n\nAvailable via: `self.run_hooks()` (engine method), `context.run_id()`, `graph.name`, `node`, `attempt`, `policy`.\n\n### 2. Add ParallelStart hook call in `parallel.rs`\n\nFile: `lib/crates/fabro-workflows/src/handler/parallel.rs`\n\nAfter `ParallelStarted` event emission (~line 301). Call `hook_runner.run()` directly since the handler has `services: &EngineServices` (which has `hook_runner` and `sandbox`):\n\n```rust\nif let Some(ref runner) = services.hook_runner {\n    let hook_ctx = HookContext::new(\n        HookEvent::ParallelStart,\n        context.run_id(),\n        graph.name.clone(),\n    );\n    let _ = runner\n        .run(&hook_ctx, services.sandbox.clone(), None)\n        .await;\n}\n```\n\n### 3. Add ParallelComplete hook call in `parallel.rs`\n\nSame file, after `ParallelCompleted` event emission (~line 713). Same pattern as ParallelStart but with `HookEvent::ParallelComplete`.\n\n### 4. Mark SandboxReady/SandboxCleanup as reserved\n\nFile: `lib/crates/fabro-workflows/src/hook/types.rs`\n\nAdd doc comments:\n```rust\n/// Reserved: hooks for this event are not yet invoked by the engine.\nSandboxReady,\n/// Reserved: hooks for this event are not yet invoked by the engine.\nSandboxCleanup,\n```\n\n### 5. Update docs\n\nFile: `docs/agents/hooks.mdx`\n\nIn the lifecycle events table, annotate `sandbox_ready` and `sandbox_cleanup` rows — e.g., append \"(not yet wired)\" to the description or add a note.\n\n## Verification\n\n1. `cargo fmt --check --all`\n2. `cargo clippy --workspace -- -D warnings`\n3. `cargo test --workspace` — existing tests pass, no regressions\n4. Manual test: create a workflow with a retrying node and a `post_tool_use`-style hook for `stage_retrying`, verify it fires in the CLI log\n\n",
+    "graph.goal": "# Wire up missing hook invocations\n\n## Context\n\nFive `HookEvent` variants exist in the enum and are documented in `docs/agents/hooks.mdx`, but `run_hooks()` is never called for them in the engine. Users can configure hooks for these events, but they silently never fire.\n\n**Events to wire up:** `StageRetrying`, `ParallelStart`, `ParallelComplete`\n**Events to mark as reserved:** `SandboxReady`, `SandboxCleanup` (sandbox lifecycle is managed outside the engine; wiring these requires significant architecture changes)\n\n## Changes\n\n### 1. Add StageRetrying hook calls in `engine.rs`\n\nFile: `lib/crates/fabro-workflows/src/engine.rs`\n\nTwo sites in `execute_with_retry`, both immediately after `WorkflowRunEvent::StageRetrying` emission and before `tokio::time::sleep(delay).await`:\n\n- **Site A (~line 1127):** error-retry path\n- **Site B (~line 1155):** explicit Retry status path\n\nPattern (same for both sites):\n```rust\n{\n    let mut hook_ctx = HookContext::new(\n        HookEvent::StageRetrying,\n        context.run_id(),\n        graph.name.clone(),\n    );\n    hook_ctx.node_id = Some(node.id.clone());\n    hook_ctx.node_label = Some(node.label().to_string());\n    hook_ctx.handler_type = node.handler_type().map(String::from);\n    hook_ctx.attempt = Some(usize::try_from(attempt).unwrap_or(usize::MAX));\n    hook_ctx.max_attempts = Some(\n        usize::try_from(policy.max_attempts).unwrap_or(usize::MAX),\n    );\n    let _ = self.run_hooks(&hook_ctx, None).await;\n}\n```\n\nAvailable via: `self.run_hooks()` (engine method), `context.run_id()`, `graph.name`, `node`, `attempt`, `policy`.\n\n### 2. Add ParallelStart hook call in `parallel.rs`\n\nFile: `lib/crates/fabro-workflows/src/handler/parallel.rs`\n\nAfter `ParallelStarted` event emission (~line 301). Call `hook_runner.run()` directly since the handler has `services: &EngineServices` (which has `hook_runner` and `sandbox`):\n\n```rust\nif let Some(ref runner) = services.hook_runner {\n    let hook_ctx = HookContext::new(\n        HookEvent::ParallelStart,\n        context.run_id(),\n        graph.name.clone(),\n    );\n    let _ = runner\n        .run(&hook_ctx, services.sandbox.clone(), None)\n        .await;\n}\n```\n\n### 3. Add ParallelComplete hook call in `parallel.rs`\n\nSame file, after `ParallelCompleted` event emission (~line 713). Same pattern as ParallelStart but with `HookEvent::ParallelComplete`.\n\n### 4. Mark SandboxReady/SandboxCleanup as reserved\n\nFile: `lib/crates/fabro-workflows/src/hook/types.rs`\n\nAdd doc comments:\n```rust\n/// Reserved: hooks for this event are not yet invoked by the engine.\nSandboxReady,\n/// Reserved: hooks for this event are not yet invoked by the engine.\nSandboxCleanup,\n```\n\n### 5. Update docs\n\nFile: `docs/agents/hooks.mdx`\n\nIn the lifecycle events table, annotate `sandbox_ready` and `sandbox_cleanup` rows — e.g., append \"(not yet wired)\" to the description or add a note.\n\n## Verification\n\n1. `cargo fmt --check --all`\n2. `cargo clippy --workspace -- -D warnings`\n3. `cargo test --workspace` — existing tests pass, no regressions\n4. Manual test: create a workflow with a retrying node and a `post_tool_use`-style hook for `stage_retrying`, verify it fires in the CLI log\n",
+    "failure_class": "",
+    "failure_signature": "",
+    "graph.model_stylesheet": "\n            * { backend: api; model: claude-opus-4-6;}\n        ",
+    "internal.fidelity": "compact",
+    "internal.retry_count.start": 1,
+    "internal.run_id": "01KKT5VGGG35Y853TNCTBVB39R",
+    "command.output": "cargo 1.94.0 (85eff7c80 2026-01-15)\n",
+    "current_node": "toolchain",
+    "command.stderr": "",
+    "internal.node_visit_count": 1,
+    "internal.thread_id": "start",
+    "graph.rankdir": "LR",
+    "outcome": "success",
+    "internal.retry_count.toolchain": 1,
+    "thread.start.current_node": "toolchain"
+  },
+  "logs": [],
+  "node_outcomes": {
+    "toolchain": {
+      "status": "success",
+      "context_updates": {
+        "command.stderr": "",
+        "command.output": "cargo 1.94.0 (85eff7c80 2026-01-15)\n"
+      },
+      "notes": "Script completed: command -v cargo >/dev/null || { curl --proto '=https' --tlsv1.2 -sSf https://sh.rustup.rs | sh -s -- -y && sudo ln -sf $HOME/.cargo/bin/* /usr/local/bin/; }; cargo --version 2>&1",
+      "duration_ms": 43
+    },
+    "start": {
+      "status": "success",
+      "duration_ms": 0
+    }
+  },
+  "next_node_id": "preflight_compile",
+  "node_visits": {
+    "start": 1,
+    "toolchain": 1
+  }
+}
\ No newline at end of file
diff --git a/nodes/start/status.json b/nodes/start/status.json
new file mode 100644
index 000000000..84f2a93d2
--- /dev/null
+++ b/nodes/start/status.json
@@ -0,0 +1,6 @@
+{
+  "status": "success",
+  "notes": null,
+  "failure_reason": null,
+  "timestamp": "2026-03-16T01:58:37.607761+00:00"
+}
\ No newline at end of file
diff --git a/nodes/toolchain/script_invocation.json b/nodes/toolchain/script_invocation.json
new file mode 100644
index 000000000..d68c414c4
--- /dev/null
+++ b/nodes/toolchain/script_invocation.json
@@ -0,0 +1,5 @@
+{
+  "command": "command -v cargo >/dev/null || { curl --proto '=https' --tlsv1.2 -sSf https://sh.rustup.rs | sh -s -- -y && sudo ln -sf $HOME/.cargo/bin/* /usr/local/bin/; }; cargo --version 2>&1",
+  "language": "shell",
+  "timeout_ms": null
+}
\ No newline at end of file
diff --git a/nodes/toolchain/script_timing.json b/nodes/toolchain/script_timing.json
new file mode 100644
index 000000000..e78aedf71
--- /dev/null
+++ b/nodes/toolchain/script_timing.json
@@ -0,0 +1,5 @@
+{
+  "duration_ms": 42,
+  "exit_code": 0,
+  "timed_out": false
+}
\ No newline at end of file
diff --git a/nodes/toolchain/status.json b/nodes/toolchain/status.json
new file mode 100644
index 000000000..c60d774d3
--- /dev/null
+++ b/nodes/toolchain/status.json
@@ -0,0 +1,6 @@
+{
+  "status": "success",
+  "notes": "Script completed: command -v cargo >/dev/null || { curl --proto '=https' --tlsv1.2 -sSf https://sh.rustup.rs | sh -s -- -y && sudo ln -sf $HOME/.cargo/bin/* /usr/local/bin/; }; cargo --version 2>&1",
+  "failure_reason": null,
+  "timestamp": "2026-03-16T01:58:37.664074+00:00"
+}
\ No newline at end of file