Document prompt routing intervention lane

Author: Your Name

docs/research/README.md

@@ -23,7 +23,13 @@ This directory groups the research tracks and branch studies that sit alongside
   - routes tasks to `minimal_prompt`, `human_structured_prompt`, or `expanded_constrained_prompt`
   - pairs with `causetrace` by selecting prompt posture first and measuring the resulting trace morphology here
 
+## Workflow Interventions
+
+- [Prompt Routing Intervention](branches/cross_project_prompt_morphology/README.md)
+  - apply-first phase is active
+  - routed tasks should be treated as a distinct prompt-posture lane in later morphology analysis
+  - use when the task framing itself is part of the experiment or workflow policy
+
 ## Boundary
 
 Branch studies and skills may inform hypotheses and workflow choices, but they do not change the `causetrace` core boundary by themselves.
-

docs/research/branches/cross_project_prompt_morphology/README.md

@@ -135,6 +135,14 @@ Detailed project-level execution records are kept outside the public `causetrace
 - [prompt-routing](https://github.com/milkoor/prompt-routing-skill) provides the prompt-posture routing skill used to apply these findings in real agent workflows.
 - It pairs with `causetrace` by selecting the prompt posture first and measuring the resulting morphology here.
 
+## Prompt-Routing Intervention Lane
+
+When this branch's findings are applied through `prompt-routing-skill`, treat the resulting sessions as a routed intervention lane:
+
+- record that prompt posture selection was applied
+- keep the routed lane separate from native direct-prompt traces
+- compare morphology before generalizing the effect to any project workflow
+
 ## Public Export Workflow
 
 Generate the public tree from a private source tree with: