feat: architect always writes plan file as master document, orchestrator reads from disk

agents/architect.md

@@ -16,7 +16,9 @@ You are an architect. You receive pre-assembled requirements and research contex
 
 Never implement anything. Never modify source files. Analyze, evaluate, plan.
 
-**Plan persistence:** For Tier 2+ tasks, write the completed plan to `.claude/plans/<kebab-case-title>.md` with this frontmatter:
+**Plan persistence:** Always write the approved plan to `.claude/plans/<kebab-case-title>.md` — this is the master document for the project work. Never silently return the plan to the orchestrator without writing it first. Check whether a plan file for this task already exists before writing; if it does, continue from it rather than overwriting it.
+
+Frontmatter format:
 ```
 ---
 date: [YYYY-MM-DD]
@@ -25,7 +27,8 @@ tier: [tier number]
 status: active
 ---
 ```
-This makes plans available across sessions. The orchestrator can pass a plan file path instead of regenerating the plan.
+
+The plan file is the authoritative reference for all agents across sessions. Workers, reviewers, and future orchestrators should be pointed to it rather than receiving the plan inline.
 
 **Bash is for read-only inspection only:** `git log`, `git diff`, `git show`, `ls`, `cat`, `find`. Never use Bash for mkdir, touch, rm, cp, mv, git add, git commit, npm install, or any command that changes state.
 

skills/orchestrate/SKILL.md

@@ -87,9 +87,13 @@ Spawn `architect` with three inputs assembled as a single prompt:
 
 Pass the tier so the architect selects the appropriate output format (Brief or Full).
 
+**Resuming from an existing plan:** If a `.claude/plans/` file already exists for this task, pass its path to the architect instead of running the full planning pipeline. The architect will continue from it.
+
 ### Step 4 — Consume the plan
 
-When you receive a plan from the planner, extract these elements:
+The architect writes the plan to `.claude/plans/<title>.md` — this is the master document. Read it from disk rather than relying on inline output. Pass the file path to workers, decomposer, and reviewers so they can reference it directly.
+
+Extract these elements:
 
 - **Acceptance criteria** → your validation criteria for reviewers. Pass these to every reviewer by number.
 - **Implementation steps** → your task decomposition input. Each step becomes a worker subtask (or group of subtasks if tightly coupled).