feat(topology): workflow of topologies (run_workflow)

Milestone-A second capability: chain whole topology runs in sequence, threading
each stage's final output into the next stage's task (e.g. swarm brainstorm →
hierarchical execute → debate review). Each stage is a full safe topology run,
so §15 holds at every step. WorkflowRecord aggregates per-stage RunRecords +
totals. Demonstrated in the benchmark example. 14 tests; clippy clean.

Co-Authored-By: Claude Opus 4.8 <[email protected]>

crates/cm-orchestrator/examples/topology_bench.rs

@@ -15,7 +15,7 @@
 use std::sync::Arc;
 
 use cm_llm::LlmProvider;
-use cm_orchestrator::{compare, ProviderExecutor, RunRecord, Scorer};
+use cm_orchestrator::{compare, run_workflow, ProviderExecutor, RunRecord, Scorer};
 use cm_topology::{Edge, EdgeKind, Node, TopologyGraph, TopologyKind};
 
 /// Deterministic quality proxy: richer (longer) output scores higher.
@@ -130,4 +130,16 @@ async fn main() {
     if let Some(i) = cmp.best_value {
         println!("best value:   {:?} (quality per token)", cmp.results[i].kind);
     }
+
+    // Workflow of topologies: brainstorm (swarm) → execute (hierarchical) → review (debate).
+    let wf = run_workflow(&[swarm(), hierarchical(), debate()], task, &executor)
+        .await
+        .expect("workflow failed");
+    println!("\nworkflow: swarm -> hierarchical -> debate");
+    println!(
+        "  stages: {}   turns: {}   tokens: {}",
+        wf.stages.len(),
+        wf.totals.turns,
+        wf.totals.tokens
+    );
 }